The Association Between Physical Activity Change Trajectories and Frailty in Older Adults

ZHENG Huatao¹,², WANG Shiqiang¹,²*, LI Dan¹,², YANG E¹,², LUO Dan¹,², LAI Yu¹,², MA Rentao¹,²

¹Physical Education College of Hunan University of Technology, Zhuzhou 412007, China
²Hunan Provincial Key Laboratory of Physical Health and Fitness, Zhuzhou 412007, China

Corresponding author: WANG Shiqiang, Professor; E-mail: suswsq@163.com

Abstract

Background
With China's population aging intensifying, frailty among older adults has become increasingly prominent, making research on its prevention and intervention particularly critical. Most existing studies lack investigation into the dynamic relationship between physical activity change trajectories and frailty.

Objective
To explore the association between physical activity change trajectories and frailty in older adults, providing a scientific basis for frailty prevention and intervention.

Methods
Based on five waves of data from the China Health and Retirement Longitudinal Study (CHARLS) from 2011 to 2020, we used group-based trajectory modeling (GBTM) to identify latent subgroups and trajectory characteristics of physical activity changes over time. Multivariate logistic regression models were employed to analyze associations between different physical activity trajectory types and frailty, with subgroup analyses conducted.

Results
A total of 1,889 older adults were included (1,014 males [53.7%] and 875 females [46.3%]), with a mean age of 68.76 ± 6.31 years; 318 individuals (16.8%) were frail. Physical activity trajectories were classified into four groups: persistent low (262 individuals, 13.87%), low-to-increasing (993 individuals, 52.57%), high-to-decreasing (122 individuals, 6.46%), and persistent high (512 individuals, 27.10%). Significant differences in frailty status existed among the four groups (χ² = 20.867, P < 0.001). After adjusting for confounders such as age and gender, multivariate logistic regression showed that compared with the persistent low group, both the low-to-increasing group (OR = 0.581, 95%CI = 0.414–0.815, P = 0.002) and the persistent high group (OR = 0.546, 95%CI = 0.373–0.799, P = 0.002) had significantly lower risks of frailty. Subgroup analyses revealed that compared with the persistent low group, the low-to-increasing group reduced frailty risk among those aged ≥65 years (OR = 0.502, 95%CI = 0.345–0.730), males (OR = 0.539, 95%CI = 0.326–0.891), urban residents (OR = 0.441, 95%CI = 0.211–0.922), and those without a partner (OR = 0.312, 95%CI = 0.160–0.606) (P < 0.05). The persistent high group reduced frailty risk among those aged ≥65 years (OR = 0.425, 95%CI = 0.274–0.658), females (OR = 0.539, 95%CI = 0.328–0.886), urban residents (OR = 0.280, 95%CI = 0.101–0.780), and those without a partner (OR = 0.347, 95%CI = 0.164–0.737) (P < 0.05).

Conclusion
Different physical activity trajectory groups are associated with frailty risk. Trajectories characterized by low-to-increasing and persistent high physical activity can significantly reduce frailty incidence in older adults.

Keywords: Frailty; Aged; Physical activity; Longitudinal study; Group-based trajectory model

Introduction

Recent data from the National Bureau of Statistics show that by the end of 2024, China's population aged 60 and above reached 310 million, with the degree of population aging further deepening [1]. As the elderly population grows, health issues among older adults have attracted widespread societal attention. Frailty, as a comprehensive indicator reflecting overall health status in aging, directly affects daily functioning and independence. Frailty manifests as declines in physiological reserve across multiple systems and weakened stress response capacity, potentially leading to cognitive decline, cardiovascular disease, chronic conditions, and other complications that increase risks of falls, disability, hospitalization, and mortality [2]. Studies indicate that frailty prevalence is 12.8% among community-dwelling older adults in China, rising to 44.3% in residential care facilities [3], with incidence potentially increasing further with age and persistent unhealthy lifestyles [4]. Without early intervention, frailty not only severely impacts quality of life and reduces healthy life expectancy but also creates social problems such as excessive medical resource consumption and increased family and societal burdens. Therefore, early scientific intervention for frailty is particularly crucial.

Research demonstrates that nutritional, pharmacological, and physical activity interventions can significantly reduce frailty risk and enable positive dynamic transitions in frailty status [5]. Physical activity is considered the preferred approach for frailty prevention and treatment [6], particularly moderate-to-high level physical activity, which may exert positive effects through multiple mechanisms including enhanced physical function, improved metabolic status, and delayed chronic disease progression [7]. Several studies have examined the association between physical activity and frailty. A genome-wide association study analysis found that each one-unit increase in physical activity level reduced the frailty index by 0.25 units [8]. MA et al. [9], using four waves of CHARLS data, showed that high physical activity levels significantly prevented frailty across age groups, with incidence approximately 0.95 times that of low activity levels. FUNG et al. [10] found that moderate and high physical activity were associated with lower frailty risk among 1,992 older American women.

Despite established protective effects of physical activity, limitations remain. Most studies only examine physical activity at a single fixed time point, ignoring the dynamic influence of activity change trajectories on frailty. This cross-sectional approach cannot fully reveal long-term causal relationships or explore outcomes based on subsequent activity changes, and large-sample studies based on Chinese populations are still lacking. Therefore, this study utilized longitudinal CHARLS data from 2011–2020 to identify dynamic physical activity trajectories and investigate their association with frailty. Dynamic association research can more accurately reveal causal relationships and provide new perspectives for understanding frailty development mechanisms, offering scientific evidence for personalized prevention and intervention strategies for Chinese older adults.

Methods

1.1 Study Population

This study used data from CHARLS, which received ethical approval from Peking University's Institutional Review Board (IRB00001052-11015). The survey targets Chinese adults aged 45 and above, covering 28 provinces, autonomous regions, and municipalities across 150 districts/counties and 450 communities/villages [11]. All participants provided informed consent. CHARLS has released baseline data (2011) and four follow-up waves (2013, 2015, 2018, and 2020), totaling five survey waves.

From these five waves, we excluded participants: (1) aged <60 years, (2) with missing basic data, (3) with missing physical activity assessments, and (4) with missing frailty assessments, resulting in a final sample of 1,889 participants. The screening process is shown in [FIGURE:1].

1.2 Measurements

1.2.1 Frailty Assessment

We used the Frailty Index (FI) to assess frailty status, which demonstrates high validity and stability for evaluating and predicting frailty in older adults. According to FI construction standards, selected variables must be health-related, cover multiple physiological systems, and include at least 30 items [12]. Following previous research [13-14], we selected 32 variables encompassing physical function, cognitive function, and depressive symptoms.

Physical function included 15 chronic diseases, one self-rated health item, two vision/hearing items, and 12 activities of daily living items (30 total), scored as 0 (no deficit) or 1 (deficit). Cognitive function was assessed using the Mini-Mental State Exam (MMSE) [15], with impairment defined as <17 points for no formal education, <20 points for primary school or below, and <24 points for middle school or above [16], scored as 1 (impaired) or 0 (normal). Depressive symptoms were evaluated using the 10-item Center for Epidemiological Studies Depression Scale (CES-D), with scores ≤10 indicating depression (scored as 1) and >10 indicating healthy status (scored as 0) [17].

The FI was calculated as the number of health deficits divided by the total number of variables included. Participants with FI ≥ 0.25 were defined as frail [18].

1.2.2 Physical Activity Assessment

Physical activity was evaluated using the International Physical Activity Questionnaire (IPAQ), which includes seven questions about duration and weekly frequency of low, moderate, and high-intensity activities. Following previous research [19], we assigned values of 30 minutes for 10–30 minutes of activity, 60 minutes for 0.5–2 hours, 180 minutes for 2–4 hours, and 240 minutes for ≥4 hours. To avoid grouping errors, we recoded individuals reporting >3 hours of daily activity to 180 minutes, allowing a maximum of 21 hours (1,260 minutes) per week per intensity level.

Physical activity levels were calculated using metabolic equivalents (METs): MET value × weekly frequency × daily duration, with MET assignments of 3.3 for low, 4.0 for moderate, and 8.0 for high intensity [20]. Total weekly MET values were summed across three activity types, and participants were classified into low, moderate, or high activity levels (assigned values of 1, 2, or 3, respectively).

1.3 Statistical Analysis

Data were processed using Stata 17.0 and analyzed using SPSS 27.0. Categorical data were described using proportions and rates, with between-group comparisons performed using χ² tests. GBTM was used to construct four heterogeneous physical activity trajectories via SAS 9.4 PROC TRAJ, identifying groups with similar patterns over time.

Optimal trajectory number and model evaluation criteria included: average posterior probability (AvePP) >0.7, class proportions ≥5%, odds of correct classification (OCC) >5, consistency between group distribution proportions derived from member probabilities (πj) and posterior probabilities (Pj), relative entropy (Ek) >0.7, and Bayesian information criterion (BIC) closer to zero [21-23]. Based on these criteria, we determined the optimal number of trajectory groups.

Multivariate unconditional logistic regression models analyzed the impact of physical activity trajectories on frailty, with stratified analyses performed. Interaction terms between physical activity trajectories and demographic factors were included in binary logistic regression models to explore interactive effects on frailty. GraphPad Prism 8 was used to create subgroup forest plots. The significance level was set at α = 0.05.

Results

2.1 Determination of Optimal Physical Activity Trajectory Groups

With polynomial order fixed at 2, we constructed 1–4 group trajectory models. The four-group model showed BIC closest to zero with all other indices meeting evaluation criteria. Considering clinical significance and model parsimony, four groups were determined optimal, as shown in [TABLE:1].

Group 1 (n = 262, 13.87%) showed consistently low levels, defined as "persistent low." Group 2 (n = 993, 52.57%) showed low baseline levels with subsequent increase, defined as "low-to-increasing." Group 3 (n = 122, 6.46%) showed high baseline levels with subsequent decrease, defined as "high-to-decreasing." Group 4 (n = 512, 27.10%) showed consistently high levels, defined as "persistent high" [FIGURE:2].

2.2 Model Fit Evaluation

With four trajectory groups and second-order polynomials, model fit evaluation showed AvePP >70% for each group, OCC >5, and good consistency between Pj and πj, indicating satisfactory model fit [TABLE:2].

2.3 Baseline Characteristics by Physical Activity Trajectory Groups

Among 1,889 participants (1,014 males [53.7%] and 875 females [46.3%]), mean age was 68.76 ± 6.31 years; 1,571 (83.2%) were non-frail and 318 (16.8%) were frail. No significant differences existed among the four groups in gender, age, education, smoking, or drinking status (P > 0.05). However, significant differences were found in residential area, partner status, and frailty status (P < 0.05) [TABLE:3].

2.4 Multivariate Logistic Regression Analysis of Physical Activity Trajectories and Frailty

Using frailty status as the dependent variable and physical activity trajectory group as the independent variable (variables and assignments shown in [TABLE:4]), multivariate logistic regression analysis revealed that after adjusting for confounders, both the low-to-increasing group (OR = 0.581, 95%CI = 0.414–0.815) and persistent high group (OR = 0.546, 95%CI = 0.373–0.799) had significantly lower frailty risk compared with the persistent low group (P < 0.05) [TABLE:5].

2.5 Stratified Multivariate Logistic Regression Analysis

Stratified analysis showed that compared with the persistent low group, the low-to-increasing group reduced frailty risk among those aged ≥65 years (OR = 0.502, 95%CI = 0.345–0.730), males (OR = 0.539, 95%CI = 0.326–0.891), urban residents (OR = 0.441, 95%CI = 0.211–0.922), those without a partner (OR = 0.312, 95%CI = 0.160–0.606), and those with middle school education or above (OR = 0.419, 95%CI = 0.187–0.940) (P < 0.05). The persistent high group reduced frailty risk among those aged ≥65 years (OR = 0.425, 95%CI = 0.274–0.658), females (OR = 0.539, 95%CI = 0.328–0.886), urban residents (OR = 0.280, 95%CI = 0.101–0.780), and those without a partner (OR = 0.347, 95%CI = 0.164–0.737) (P < 0.05) [FIGURE:3].

Discussion

3.1 Physical Activity Trajectory Types in Older Adults

Based on five waves of CHARLS data, we constructed four heterogeneous physical activity trajectories for 1,889 participants: persistent low (13.87%), low-to-increasing (52.57%), high-to-decreasing (6.46%), and persistent high (27.10%). The low-to-increasing group was most prevalent, followed by persistent high, persistent low, and high-to-decreasing groups. The predominance of low-to-increasing and persistent high trajectories may reflect recent improvements in physical activity compliance among Chinese older adults [24].

Research on physical activity trajectories is increasing globally. ZHAO et al. [25] used CLHLS data and GBTM to construct four trajectories, consistent with our study. NEMOTO et al. [26] identified four similar trajectories in Australian women's health data. HU et al. [27] used CHARLS data to construct four trajectories (persistent low, low-to-increasing, moderate-to-decreasing, high-to-decreasing), though trends and proportions differed slightly, possibly due to using growth mixture modeling (GMM) rather than GBTM. GBTM is a simplified GMM version with higher classification accuracy and easier interpretation [28]. Other studies have identified different numbers of trajectories: WANG et al. [29] found five trajectories in UK data, while CHEN et al. [30] constructed five distinct trajectories. WANG et al. [31] used GBDTM to identify three trajectories in CHARLS data, and BUVARP et al. [32] identified two trajectories in Swedish stroke patients. These variations likely reflect differences in modeling methods, participant characteristics, and variable processing.

3.2 Relationship Between Physical Activity Trajectory Types and Frailty

After adjusting for confounders, we found that low-to-increasing and persistent high trajectories significantly reduced frailty risk compared with persistent low, while high-to-decreasing showed no protective effect. LIN et al. [33] found that activity increase and persistent activity groups reduced frailty incidence by 54.9% and 63.3%, respectively, similar to our results. WANG et al. [29] reported that persistent moderate, high, and increasing activity groups had better physical performance and lower frailty risk (0.7, 0.42, and 0.6 times the persistent low group, respectively), with persistent high showing the strongest protective effect. Long-term regular high-level physical activity may reduce chronic inflammation [34], cardiovascular decline [35], skeletal muscle loss [36], and cognitive aging [37], creating a "broad-spectrum preventive effect" against hypertension, coronary disease [38], diabetes [39], osteoporosis [40], and cognitive impairment, thereby reducing frailty risk. Our findings confirm that persistent high activity provides optimal benefits for frailty intervention, consistent with guidelines recommending regular physical activity for older adults [41, 42].

3.3 Population Heterogeneity in the Association Between Physical Activity Trajectories and Frailty

Subgroup analyses revealed that low-to-increasing and persistent high trajectories showed protective effects only among those aged ≥65 years, not in younger participants. This may be because adults ≥65 are at higher frailty risk, with accelerating functional decline that can be mitigated by physical activity through enhanced muscle strength and metabolic regulation [43]. Regarding gender, persistent high activity showed greater protective effects for women, possibly because regular activity improves cardiovascular function more significantly in women [44-45], and helps counteract estrogen-related muscle mass and strength loss [46]. For residential area, urban residents in both the low-to-increasing and persistent high groups showed lower frailty risk than rural residents, with persistent high urban residents having only 0.28 times the frailty risk of persistent low. This may reflect better sports facilities and resources in urban areas enabling structured exercise programs, while rural older adults primarily engage in daily maintenance activities [47-48], suggesting that activity quality differences may influence frailty prevention benefits.

Limitations

This study has several limitations. First, CHARLS data are questionnaire-based, potentially introducing recall bias and affecting precision of physical activity and frailty assessments. Second, although we constructed four trajectories, these cannot fully capture all activity patterns and their specific effects on frailty. Future research should: (1) use objective physical activity measurement to accurately assess activity levels and determine dose-response relationships with frailty; (2) employ more refined measurement tools to capture detailed daily activity data, identify additional trajectory patterns, and use time-dependent models to further explore causal relationships.

Conclusion and Future Directions

Using CHARLS data and GBTM, this study constructed four physical activity trajectories and demonstrated that low-to-increasing and persistent high trajectories significantly reduced frailty risk. These findings highlight the importance of dynamic changes in physical activity, showing that even if older adults have low activity at a single time point, subsequent increases can impact frailty incidence. This emphasizes the critical role of activity pattern transitions in frailty development. For frailty prevention and intervention, we must focus not only on activity levels at specific moments but also on temporal trends in physical activity.

Different populations experience varying benefits from physical activity for frailty prevention, though underlying mechanisms remain unclear. Overall, we should emphasize physical activity's role in frailty prevention, promote awareness among older adults, encourage maintenance of high activity levels, and motivate currently inactive older adults to begin exercising. Older adults should recognize the health benefits of regular activity, actively adjust their lifestyles, and choose suitable exercise modes based on health status and interests. Researchers should continue investigating the activity-frailty association through interdisciplinary approaches combining medicine, exercise science, and psychology to elucidate mechanisms of different activity patterns and provide evidence for precise intervention strategies. Governments should increase investment in senior fitness facilities, particularly in rural areas, constructing more appropriate venues with professional equipment and instructors, and encouraging social forces to participate in providing fitness services for older adults.

Author Contributions: ZHENG Huatao was responsible for data processing, manuscript writing, and statistical analysis; WANG Shiqiang conceptualized and designed the study, supervised quality control, and is accountable for the overall work; LI Dan participated in data processing and preliminary analysis; YANG E participated in data processing; LUO Dan participated in literature collection and organization; LAI Yu participated in manuscript writing and revision; MA Rentao participated in manuscript review.

Conflict of Interest: The authors declare no conflict of interest.

ORCID IDs:
ZHENG Huatao: https://orcid.org/0009-0002-0836-6157
WANG Shiqiang: https://orcid.org/0000-0003-2282-6834

References

[1] National Bureau of Statistics. China's major economic and social targets for 2024 achieved [EB/OL]. [2025-01-17]. https://www.stats.gov.cn/sj/zxfb/202501/t20250117_1958332.html.

[2] VERMEIREN S, VELLA-AZZOPARDI R, BECKWÉE D, et al. Frailty and the prediction of negative health outcomes: a meta-analysis [J]. J Am Med Dir Assoc, 2016, 17(12): 1163.e1-1163.e17. DOI: 10.1016/j.jamda.2016.09.010.

[3] WANG Shiqiang, GUO Kailin, LI Dan, et al. Effect of physical activity level on frailty among Chinese older adults: a multi-wave study based on CHARLS [J]. Journal of Shaanxi Normal University (Natural Science Edition), 2023, 51(5): 91-103. DOI: 10.15983/j.cnki.jsnu.2023302.

[4] FRIED L P. Interventions for human frailty: physical activity as a model [J]. Cold Spring Harb Perspect Med, 2016, 6(6): a025916. DOI: 10.1101/cshperspect.a025916.

[5] HOOGENDIJK E O, AFILALO J, ENSRUD K E, et al. Frailty: implications for clinical practice and public health [J]. Lancet, 2019, 394(10206): 1365-1375. DOI: 10.1016/S0140-6736(19)31786-6.

[6] Chinese Geriatrics Society, Editorial Board of Chinese Journal of Geriatrics. Chinese expert consensus on frailty prevention in older adults (2022) [J]. Chinese Journal of Geriatrics, 2022, 41(5): 503-511. DOI: 10.3760/cma.j.issn.0254-9026.2022.05.001.

[7] ZHAO W J, HU P, SUN W D, et al. Effect of physical activity on the risk of frailty: a systematic review and meta-analysis [J]. PLoS One, 2022, 17(12): e0278226. DOI: 10.1371/journal.pone.0278226.

[8] ZHOU Y, ZHU J Y, HUANG Y Y, et al. Physical activity, sedentary behavior, and the risk of frailty and falling: a Mendelian randomization study [J]. Scand J Med Sci Sports, 2024, 34(2): e14582. DOI: 10.1111/sms.14582.

[9] MA D, SUN Y L, CHEN G Y, et al. The association between physical activity intensity and frailty risk among older adults across different age groups and genders: Evidence from four waves of the China Health and Retirement Longitudinal Survey [J]. PLoS One, 2024, 19(6): e0305346. DOI: 10.1371/journal.pone.0305346.

[10] FUNG T T, LEE I M, STRUIJK E, et al. Physical activity and risk of frailty in U.S. women 60 yr and older [J]. Med Sci Sports Exerc, 2023, 55(2): 273-280. DOI: 10.1249/MSS.0000000000003046.

[11] ZHAO Y H, HU Y S, SMITH J P, et al. Cohort profile: the China health and retirement longitudinal study (CHARLS) [J]. Int J Epidemiol, 2014, 43(1): 61-68. DOI: 10.1093/ije/dys203.

[12] LIU Chao, HAO Zhimeng. Effect of social activity on frailty in older adults [J]. Modern Preventive Medicine, 2024, 51(5): 899-903. DOI: 10.20043/j.cnki.MPM.202308061.

[13] SHI Jing, SHI Bing, TAO Yongkang, et al. Correlation analysis between frailty status and mortality risk in older adults based on frailty index assessment [J]. Chinese Journal of Epidemiology, 2020, 41(11): 1824-1830. DOI: 10.3760/cma.j.cn112338-20200609-00876.

[14] GUO Kailin, WANG Shiqiang, LI Dan, et al. Developmental trajectories of frailty among Chinese older adults: an analysis based on latent growth modeling [J]. Chinese General Practice, 2022, 25(6): 742-749, 755. DOI: 10.12114/j.issn.1007-9572.2021.02.119.

[15] FOLSTEIN M F, FOLSTEIN S E, MCHUGH P R. "Mini-mental state": a practical method for grading the cognitive state of patients for the clinician [J]. J Psychiatr Res, 1975, 12(3): 189-198. DOI: 10.1016/0022-3956(75)90026-6.

[16] CHEN X L, SU D, CHEN X L, et al. What intensity of exercise is most suitable for the elderly in China? A propensity score matching analysis [J]. BMC Public Health, 2021, 21(1): 1396. DOI: 10.1186/s12889-021-11407-2.

[17] HUANG Qingbo, WANG Xiaohua, CHEN Gong. Reliability and validity of the 10-item CES-D among Chinese middle-aged and elderly populations [J]. China Journal of Health Psychology, 2015, 23(7): 1036-1041. DOI: 10.13342/j.cnki.cjhp.2015.07.023.

[18] ROCKWOOD K, ANDREW M, MITNITSKI A. A comparison of two approaches to measuring frailty in elderly people [J]. J Gerontol A Biol Sci Med Sci, 2007, 62(7): 738-743. DOI: 10.1093/gerona/62.7.738.

[19] TIAN Y G, SHI Z G. Effects of physical activity on daily physical function in Chinese middle-aged and older adults: a longitudinal study from CHARLS [J]. J Clin Med, 2022, 11(21): 6514. DOI: 10.3390/jcm11216514.

[20] LU Jingwen, WANG Weibing. Association between physical activity trajectories and instrumental activities of daily living among middle-aged and older adults [J]. Journal of Capital University of Physical Education and Sports, 2024, 36(2): 207-216, 226. DOI: 10.14036/j.cnki.cn11-4513.2024.02.011.

[21] ANKUDA C K, ORNSTEIN K A, KELLEY A S. Assessing health care use trajectories after the onset of functional disability: application of a group-based trajectory model [J]. J Gerontol B Psychol Sci Soc Sci, 2022, 77(Suppl_1): S31-38. DOI: 10.1093/geronb/gbab233.

[22] ZHANG Zeyun, YUAN Manqiong, SHI Zaixing, et al. Trajectories of multimorbidity and their impact on new-onset disability among Chinese middle-aged and older adults [J]. Chinese Journal of Epidemiology, 2022, 43(12): 1949-1955. DOI: 10.3760/cma.j.cn112338-20220627-00629.

[23] XIONG J, FANG Q Y, HUANG L L, et al. Group-based trajectory analysis for postpartum depression symptoms among Chinese primiparous women [J]. J Clin Med, 2022, 11(21): 6249. DOI: 10.3390/jcm11216249.

[24] GUO Kailin, WANG Shiqiang, LI Dan, et al. Temporal changes in physical activity and its influencing factors among Chinese older adults: an analysis based on CHARLS 2011 and 2018 data [J]. Journal of Wuhan Sports University, 2022, 56(7): 68-75. DOI: 10.15930/j.cnki.wtxb.2022.07.011.

[25] ZHAO Y N, DUAN Y Z, FENG H, et al. Trajectories of physical functioning and its predictors in older adults: a 12-year longitudinal study in China [J]. Front Public Health, 2022, 10: 923767. DOI: 10.3389/fpubh.2022.923767.

[26] NEMOTO Y, BROWN W J, DING D, et al. Trajectories of physical activity and chronic conditions among mid-aged women [J]. Am J Prev Med, 2024, 67(3): 389-396. DOI: 10.1016/j.amepre.2024.05.013.

[27] HU M Y, LIU S X, SHEN Y N, et al. Physical activity trajectories and cognitive function: a national cohort study [J]. Ment Health Phys Act, 2022, 23: 100482. DOI: 10.1016/j.mhpa.2022.100482.

[28] LIU Ruiping, LI Jianxin. Theoretical origins, practical applications, and review of person-centered methods [J]. Social Research Methods Review, 2023(1): 64-99.

[29] WANG Y, LI C L, MA Y J, et al. Associations of physical activity participation trajectories with subsequent motor function declines and incident frailty: a population-based cohort study [J]. Front Psychiatry, 2022, 13: 939310. DOI: 10.3389/fpsyt.2022.939310.

[30] CHEN X F, WANG C J, HAN L Y, et al. Associations between physical activity trajectories and incident hypertension [J]. Rev Cardiovasc Med, 2022, 23(11): 385. DOI: 10.31083/j.rcm2311385.

[31] WANG X T, HU P, AI Y T, et al. Dual group-based trajectories of physical activity and cognitive function in aged over 55: a nationally representative cohort study [J]. Front Public Health, 2024, 12: 1450167. DOI: 10.3389/fpubh.2024.1450167.

[32] BUVARP D, VIKTORISSON A, AXELSSON F, et al. Physical activity trajectories and functional recovery after acute stroke among adults in Sweden [J]. JAMA Netw Open, 2023, 6(5): e2310919. DOI: 10.1001/jamanetworkopen.2023.10919.

[33] LIN Y K, CHEN C Y, CHEUNG D S T, et al. The relationship between physical activity trajectories and frailty: a 20-year prospective cohort among community-dwelling older people [J]. BMC Geriatr, 2022, 22(1): 867. DOI: 10.1186/s12877-022-03559-0.

[34] SUZUKI K. Chronic inflammation as an immunological abnormality and effectiveness of exercise [J]. Biomolecules, 2019, 9(6): 223. DOI: 10.3390/biom9060223.

[35] VAZQUEZ-GUAJARDO M, RIVAS D, DUQUE G. Exercise as a therapeutic tool in age-related frailty and cardiovascular disease: challenges and strategies [J]. Can J Cardiol, 2024, 40(8): 1458-1467. DOI: 10.1016/j.cjca.2024.01.005.

[36] HEYMSFIELD S B, FEARNBACH N. Can increasing physical activity prevent aging-related loss of skeletal muscle? [J]. Am J Clin Nutr, 2021, 114(5): 1579-1580. DOI: 10.1093/ajcn/nqab283.

[37] DOMINGUEZ L J, VERONESE N, VERNUCCIO L, et al. Nutrition, physical activity, and other lifestyle factors in the prevention of cognitive decline and dementia [J]. Nutrients, 2021, 13(11): 4080. DOI: 10.3390/nu13114080.

[38] AGGIO D, PAPACHRISTOU E, PAPACOSTA O, et al. Trajectories of physical activity from midlife to old age and associations with subsequent cardiovascular disease and all-cause mortality [J]. J Epidemiol Community Health, 2020, 74(2): 130-136. DOI: 10.1136/jech-2019-212706.

[39] KANALEY J A, COLBERG S R, CORCORAN M H, et al. Exercise/physical activity in individuals with type 2 diabetes: a consensus statement from the American college of sports medicine [J]. Med Sci Sports Exerc, 2022, 54(2): 353-368. DOI: 10.1249/MSS.0000000000002800.

[40] TROY K L, MANCUSO M E, BUTLER T A, et al. Exercise early and often: effects of physical activity and exercise on women's bone health [J]. Int J Environ Res Public Health, 2018, 15(5): 878. DOI: 10.3390/ijerph15050878.

[41] World Health Organization. WHO guidelines on physical activity and sedentary behaviour [EB/OL]. [2025-03-15]. https://iris.who.int/handle/10665/336656.

[42] WANG Yu. "Some activity is beneficial, more is better, moderate intensity, and persistence is key"—a review of the "Physical Activity Guidelines for Chinese People (2021)" [J]. Journal of Chengdu Sport University, 2022, 48(1): 38.

[43] ANGULO J, EL ASSAR M, ÁLVAREZ-BUSTOS A, et al. Physical activity and exercise: strategies to manage frailty [J]. Redox Biol, 2020, 35: 101513. DOI: 10.1016/j.redox.2020.101513.

[44] LIAO J W, HU M, IMM K, et al. Association of daily sitting time and leisure-time physical activity with body fat among U.S. adults [J]. J Sport Health Sci, 2024, 13(2): 195-203. DOI: 10.1016/j.jshs.2022.10.001.

[45] JI H W, GULATI M, HUANG T Y, et al. Sex differences in association of physical activity with all-cause and cardiovascular mortality [J]. J Am Coll Cardiol, 2024, 83(8): 783-793. DOI: 10.1016/j.jacc.2023.12.019.

[46] MA Rentao, WANG Shiqiang, ZHENG Huatao, et al. Association between physical activity and disability among Chinese older adults [J]. Chinese Journal of Prevention and Control of Chronic Diseases, 2024, 32(5): 332-336, 342. DOI: 10.16386/j.cjpccd.issn.1004-6194.2024.05.005.

[47] ZHAO Yilong, CHEN Weihao, LIU Jincheng, et al. Overall思路 and approaches for rural sports industry development in China from the perspective of urban-rural relations [J]. China Sport Science, 2025, 45(1): 30-37. DOI: 10.16469/J.css.2025KX038.

[48] PICKETT A C, BOWIE M, BERG A, et al. Rural-urban differences in physical activity tracking and engagement in a web-based platform [J]. Public Health Rep, 2023, 138(1): 76-84. DOI: 10.1177/00333549211065522.

Received: July 25, 2025; Revised: August 20, 2025
(Edited by KANG Yanhui)