A Reverse J-Shaped Association Between Total Physical Activity and Cognitive Impairment in Older Chinese Adults: Evidence from a Nationally Representative Cross-sectional Study Using CHARLS Data

A Reverse J-Shaped Association Between Total Physical Activity and Cognitive Impairment in Older Chinese Adults: Evidence from a Nationally Representative Cross-sectional Study Using CHARLS Data | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article A Reverse J-Shaped Association Between Total Physical Activity and Cognitive Impairment in Older Chinese Adults: Evidence from a Nationally Representative Cross-sectional Study Using CHARLS Data Yongheng Zhao, Gaixia Hou, Limeng Liu, Lizhen Ning, Xuefeng Xi This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7282499/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background: Cognitive impairment (CI) is a growing concern in aging societies, particularly in low- and middle-income countries. Although physical activity (PA) is widely recognized as neuroprotective, its optimal dose for cognitive health remains uncertain. Objective : To evaluate the nonlinear association between total physical activity (TPA) and cognitive impairment in older Chinese adults, and to examine whether depressive symptoms and life satisfaction mediate this relationship. Methods: We analyzed data from 7,818 adults aged ≥60 years in the China Health and Retirement Longitudinal Study (CHARLS, 2011–2020). Total physical activity (TPA) was assessed using the long-form IPAQ and expressed in MET-minutes/week. CI was defined as a total cognitive score <10. Logistic regression, restricted cubic spline models, and structural equation modeling (SEM) were used to examine nonlinear associations and psychological mediation pathways, adjusting for demographic and health covariates. Results: We observed a reverse J-shaped association between TPA and CI risk, with the lowest odds at ~2,800 MET-min/week (OR = 0.772, 95% CI: 0.648–0.917). Excessive TPA was not associated with additional benefits and showed a trend toward increased risk. Mediation analysis revealed that depressive symptoms and life satisfaction partially accounted for the relationship. Conclusion: This study observed a statistically significant reverse J-shaped association between total physical activity and cognitive impairment in older Chinese adults, with the lowest risk occurring around 2,800 MET-minutes per week. Both insufficient and excessive activity levels were associated with higher cognitive risk, extending previous findings by quantifying this population's optimal activity range for cognitive health. Additionally, depressive symptoms and life satisfaction were identified as partial mediators, suggesting that psychological well-being may play a significant role in the pathway linking physical activity to cognitive outcomes among older adults. Cognitive impairment Physical activity Dose–response relationship Older adults CHARLS Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Cognitive impairment (CI) is increasingly recognized as a critical public health challenge in China, especially amid rapid population aging. Recent epidemiological data estimate that approximately 15.5% of older Chinese adults are affected by mild cognitive impairment (MCI), and 6.0% live with dementia, reflecting a substantial and growing burden nationwide [ 1 ] . The etiology of CI is multifactorial, with strong associations to modifiable cardiometabolic risk factors such as hypertension, diabetes, obesity, and depression [ 2 – 5 ] . While China has achieved a decline in age-standardized stroke incidence over the past decades, the absolute burden of cerebrovascular disease and associated cognitive sequelae continues to rise, driven by demographic transitions and expanding life expectancy [ 6 – 8 ] . The Global Burden of Disease (GBD) Study further indicates that China ranks among the countries with the highest absolute number of stroke- and dementia-related disability-adjusted life years (DALYs) globally [ 8 – 10 ] . These trends highlight the urgent need for scalable, low-cost, and preventive strategies to mitigate cognitive decline, particularly among older adults at high risk. Physical activity (PA) is a well-established, modifiable protective factor for cardiometabolic health and brain function [ 11 , 12 ] . International guidelines, such as those from the WHO, recommend that older adults perform at least 150 minutes of moderate-intensity activity (equivalent to ~ 600 MET-minutes/week) to maintain overall health [ 13 ] . Emerging neuroimaging and experimental research shows that PA promotes neurogenesis, preserves brain volume, and improves cognitive function in older adults [ 14 , 15 ] . PA has been listed among the key targets for dementia prevention by the Lancet Commission [ 16 ] . Nevertheless, the shape and strength of the association between PA and cognitive outcomes remain debated. Some studies suggest that moderate PA levels yield maximal cognitive benefit, while excessive or high-intensity physical activity may offer diminishing or adverse effects [ 17 , 18 ] . Additionally, subgroup analyses indicate that the relationship may vary by sex, education, and comorbid status [ 19 ] . Moreover, cultural differences, such as rural–urban divides or education-based disparities, may moderate these effects, particularly in middle-income settings like China [ 20 ] . This study investigated the association between total physical activity (TPA, measured in MET-minutes/week) and cognitive impairment risk in a large, nationally representative sample of Chinese older adults using the CHARLS data (2011–2020). We examined linear and non-linear dose–response patterns and evaluated effect modification by demographic and clinical subgroups. These findings aim to inform evidence-based PA recommendations to prevent cognitive decline among older Chinese adults. Materials and methods Study Design and Population This study adopted a pooled cross-sectional design based on data from the China Health and Retirement Longitudinal Study (CHARLS), a nationally representative cohort of Chinese adults aged 45 years and above. CHARLS employed a multistage, stratified, probability-proportional-to-size (PPS) sampling method covering 28 provinces, 150 counties, and 450 communities [21,22] . The baseline wave was initiated in 2011–2012, followed by follow-up assessments in 2013, 2015, 2018, and 2020. Data were collected through face-to-face computer-assisted personal interviews (CAPI). For this analysis, individuals aged 60 or older at any wave were included, yielding an initial sample of 15,264 participants. Each participant was included based on the first wave in which they met the age eligibility criterion to avoid repeated measures. Participants with missing values for key variables—including total cognition score, total physical activity (MET-min/week), sex, education, marital status, number of chronic conditions, or CESD-10 depression score—were excluded. After applying these criteria, the final complete-case sample consisted of 7,818 participants. A binary cognitive status variable (CI_group) was created, with participants classified as cognitively normal (score ≥10; n = 5,948) or cognitively impaired (score <10; n = 1,870) based on total cognition score. A detailed flow diagram outlining the sample selection process is presented in Figure 1 . Assessment of Physical Activity and Cognitive Function Assessment of Physical Activity Total physical activity (TPA) was measured using the long-form International Physical Activity Questionnaire (IPAQ-LF), a validated instrument widely used in epidemiological studies, which records the frequency and duration of physical activity across three domains: occupational, transportation, and leisure-time. The IPAQ-LF was not developed for this study, and its official development and scoring guidelines are publicly available from the IPAQ Research Committee (IPAQ Research Committee. Guidelines for data processing and analysis of the International Physical Activity Questionnaire (IPAQ) – Short and Long Forms. Revised November 2005. Available from: https://sites.google.com/view/ipaq/score). Participants self-reported the weekly time spent on walking, moderate-intensity, and vigorous-intensity activities. Activity durations were converted into metabolic equivalent task (MET) values using standard coefficients: 3.3 for walking, 4.0 for moderate-intensity activity, and 8.0 for vigorous-intensity activity [23] . Total weekly MET-minutes were computed by summing the MET values across all activity domains. TPA was analyzed in both continuous and categorical forms. TPA was modeled per 1,000 MET-minutes/week in continuous models to facilitate interpretation. For categorical analyses, participants were divided into quartiles based on weekly TPA: Q1 (≤1,611 MET-min/week), Q2 (1,612–3,489), Q3 (3,490–6,875), and Q4 (≥6,876). This approach examined linear and nonlinear dose–response relationships between TPA and cognitive outcomes. Assessment of Cognitive Function Cognitive function was evaluated using a composite score based on the cognitive assessments administered in CHARLS. This score encompassed three domains: (1) episodic memory, assessed via immediate and delayed word recall (range: 0–10); (2) orientation and attention, assessed using the TICS battery and serial subtraction tasks (range: 0–10); and (3) visuospatial ability, evaluated through a figure-drawing test (range: 0–1) [24] . A total cognitive score ranging from 0 to 21 was calculated by summing the scores across all three domains. Consistent with prior CHARLS-based studies [25] , cognitive impairment was defined as a total cognitive score <10. A binary outcome variable (CI_group) was generated, where 0 indicated cognitively normal (score ≥10) and 1 indicated cognitive impairment (score <10). Covariates Covariates were selected based on theoretical relevance and prior evidence regarding cognitive function in older adults. Sociodemographic covariates included age (continuous), sex (male or female), years of education, and marital status (married vs. unmarried). Health-related covariates included the number of self-reported chronic diseases (range: 0–10) and depressive symptoms, assessed using the 10-item Center for Epidemiologic Studies Depression Scale (CESD-10; range: 0–30), a validated short form of the CES-D originally developed by Andresen et al [26] .. All covariates were entered as control variables in the multivariable logistic regression and structural equation models. Statistical Analysis All analyses were conducted using R software (version 4.2.2). A retrospective pseudo–cross-sectional approach was adopted by selecting the most recent wave available for each participant (2011–2020) from the longitudinal CHARLS dataset, ensuring individual-level independence and eliminating repeated measures. Group differences in baseline characteristics (cognitively normal vs. impaired) were evaluated using independent-samples t-tests for normally distributed continuous variables, Mann–Whitney U tests for non-normally distributed variables, and chi-square tests for categorical variables. Effect sizes were reported as Cohen’s d for constant variables and Cramér’s V for categorical comparisons. To examine the association between total physical activity (TPA) and cognitive impairment, a series of multivariable logistic regression models were fitted as follows: • Model 1: unadjusted; • Model 2: adjusted for age, sex, and education level; • Model 3: additionally adjusted for marital status, number of chronic diseases, depressive symptoms (CES-D10 score), body mass index (BMI), smoking status, and alcohol consumption; • Model 4: TPA modeled as a continuous variable (per 1,000 MET-min/week), with the same covariates as in Model 3. Dose–response relationships were modeled using restricted cubic splines (RCS) via the rms package. Knots were placed at the 5th, 35th, 65th, and 95th percentiles of TPA, by Harrell’s guidelines. The median of the lowest TPA quartile was used as the reference point. Nonlinearity was assessed by likelihood ratio tests comparing the spline model to a nested linear model. Model fit was assessed using Nagelkerke’s pseudo R² and the Hosmer–Lemeshow goodness-of-fit test. Multicollinearity was examined using variance inflation factors (VIF), with all included covariates exhibiting VIFs less than 2.0. All statistical tests were two-sided; a p-value < 0.05 was considered statistically significant. Stratified and Interaction Analyses To assess potential heterogeneity in the association between total physical activity (TPA) and cognitive impairment across key subpopulations, stratified logistic regression analyses were conducted by sex (male vs. female), hypertension (yes vs. no), diabetes (yes vs. no), and education level (≤2 vs. ≥3 years), based on the fully adjusted model (Model 3). The same set of covariates was retained across all subgroup models to ensure comparability. Statistical interactions between TPA and each stratification variable were tested by including multiplicative interaction terms in the fully adjusted model. Interaction significance was evaluated using Wald chi-square tests, with a significance level set at p < 0.05. To account for family-wise error rate (FWER) in multiple subgroup comparisons, post hoc pairwise tests across TPA quartiles within each stratum were adjusted using the Tukey Honestly Significant Difference (HSD) method. Additionally, restricted cubic spline (RCS) modeling was applied within each subgroup to estimate predicted probabilities of cognitive impairment across the TPA continuum. Although graphical results are not shown, this approach enabled exploration of subgroup-specific dose–response relationships, allowing for a more nuanced understanding of how physical activity may differentially influence cognitive outcomes across population segments. Results Baseline Characteristics of Participants A total of 7,818 participants aged 60 years and older were included in the final analysis. Among them, 1,870 individuals (23.9%) were classified as having cognitive impairment (CI group = 1), and 5,948 (76.1%) were cognitively normal (CI group = 0), based on the total cognition score threshold of <10 points. As shown in Table 1, participants in the cognitively impaired group were significantly older than those in the cognitively normal group (mean age: 69.99 ± 6.90 vs. 68.32 ± 5.90 years, p < 0.001). They also had fewer years of education (1.55 ± 0.84 vs. 2.38 ± 1.08, p < 0.001), a lower proportion of being married (76.3% vs. 84.7%, p < 0.001), and higher levels of depressive symptoms as measured by the CES-D10 scale (10.68 ± 6.75 vs. 7.70 ± 6.00, p < 0.001). Interestingly, the cognitively impaired group reported higher total physical activity levels (5,709.56 ± 5,857.79 vs. 5,163.29 ± 5,265.74 MET-min/week, p < 0.001), suggesting a potential non-linear association between physical activity and cognitive status. Table 1. Baseline Characteristics of Participants by Cognitive Status (n = 7,818) Variable Cognitively Normal (n = 5,948) Cognitively Impaired (n = 1,870) p-value Age, years (mean ± SD) 68.32 ± 5.90 69.99 ± 6.90 <0.001 *** Male, n (%) 3,580 (60.2%) 891 (47.6%) <0.001 *** Female, n (%) 2,368 (39.8%) 981 (52.4%) <0.001 *** Years of education (mean ± SD) 2.38 ± 1.08 1.55 ± 0.84 <0.001 *** Married, n (%) 5,036 (84.7%) 1,427 (76.3%) <0.001 *** CES-D10 score (mean ± SD) 7.70 ± 6.00 10.68 ± 6.75 <0.001 *** Total PA (MET-min/week, mean ± SD) 5,163.29 ± 5,265.74 5,709.56 ± 5,857.79 <0.001 *** Note : Values are presented as mean ± standard deviation for continuous variables and number (percentage) for categorical variables. The CI group was defined based on total cognitive score: cognitively impaired (<10 points), cognitively normal (≥10 points). P-values were derived from independent t -tests (continuous variables) or chi-square tests (categorical variables). TPA = Total Physical Activity; CESD-10 = 10-item Center for Epidemiologic Studies Depression Scale. P-values indicate statistical significance levels: ***p < 0.001; **p < 0.01; *p < 0.05. Association Between Total Physical Activity and Cognitive Impairment We examined the association between total physical activity (TPA) and cognitive impairment (CI) using multivariable logistic regression models across TPA quartiles (Q1–Q4), with Q1 (<1,200 MET-min/week) serving as the reference group. As shown in Table 2, in the unadjusted model (Model 1), participants in the second quartile (Q2: ~1,200–2,800 MET-min/week) had 32.5% lower odds of cognitive impairment compared to those in Q1 (OR = 0.675, 95% CI: 0.574–0.791, p < 0.001). The third quartile (Q3: ~2,800–5,000 MET-min/week) was associated with a modest but statistically significant risk reduction (OR = 0.865, 95% CI: 0.748–0.998, p = 0.048), while those in the highest quartile (Q4: ≥5,000 MET-min/week) had a 22.3% increased risk (OR = 1.220, 95% CI: 1.070–1.390, p = 0.003). In the demographic-adjusted model (Model 2), adjusting for age, gender, education, and marital status, the protective effect of moderate TPA remained significant. Compared to Q1, Q2 (OR = 0.762, 95% CI: 0.641–0.903, p = 0.002) and Q3 (OR = 0.847, 95% CI: 0.726–0.987, p = 0.035) were associated with significantly reduced odds of CI. Conversely, Q4 was associated with increased risk (OR = 1.170, 95% CI: 1.010–1.350, p = 0.036). In the fully adjusted model (Model 3), which further controlled for chronic diseases and depressive symptoms (CESD-10), the second quartile remained significantly protective (OR = 0.772, 95% CI: 0.648–0.917, p = 0.003). The estimates for Q3 and Q4 indicated a potential reversal of effect, but did not reach statistical significance (Q3: OR = 0.864, 95% CI: 0.739–1.010, p = 0.067; Q4: OR = 1.150, 95% CI: 0.996–1.330, p = 0.057). These findings suggest a reverse J-shaped association between total physical activity and the odds of cognitive impairment. Restricted cubic spline modeling (Figure 2) further confirmed this nonlinear trend, identifying the lowest risk point at approximately 2,800 MET-min/week. Table 2. Association Between Total Physical Activity and Cognitive Impairment Among Older Adults: Multivariable Logistic Regression Models (n = 7,818) TPA Quartile Model 1 (Unadjusted) OR (95% CI) p -value Model 2 (Demographic-adjusted) OR (95% CI) p -value Model 3 (Fully-adjusted) OR (95% CI) p -value Q1 (ref) 1.00 — 1.00 — 1.00 — Q2 0.675 (0.574–0.791) <0.001 *** 0.762 (0.641–0.903) 0.002 ** 0.772 (0.648–0.917) 0.003 ** Q3 0.865 (0.748–0.998) 0.048 * 0.847 (0.726–0.987) 0.035 * 0.864 (0.739–1.010) 0.067 Q4 1.22 (1.07–1.39) 0.003 ** 1.17 (1.01–1.35) 0.036 * 1.15 (0.996–1.33) 0.057 Note: Odds ratios (ORs) and 95% confidence intervals (CIs) were estimated using logistic regression models, with cognitive impairment (CI_group) as the binary outcome (0 = cognitively normal, 1 = cognitively impaired). Total physical activity (TPA) was categorized into quartiles (Q1–Q4) based on weekly MET-minutes and modeled as a categorical variable, with Q1 (≤1,611 MET-min/week) as the reference group. Model 1 : Unadjusted Model 2 : Adjusted for age, sex, education level, and marital status Model 3 : Fully adjusted for Model 2 covariates plus number of chronic conditions and CESD-10 depressive symptoms score. P-values indicate statistical significance levels: ***p < 0.001; **p < 0.01; *p < 0.05. Estimates in bold indicate statistically significant associations at the p < 0.05 threshold. Nonlinear Dose–Response Relationship Between Total Physical Activity and Cognitive Impairment To further explore the shape of the association between total physical activity (TPA) and cognitive impairment, we applied a restricted cubic spline (RCS) regression model with four knots, adjusting for age, gender, education, marital status, chronic disease status, and depressive symptoms. The resulting dose–response curve revealed a nonlinear reverse J-shaped association between TPA and the odds of cognitive impairment (Figure 2). Specifically, the risk of cognitive impairment declined sharply with increasing TPA levels up to approximately 2,800 MET-min/week, after which the odds began to rise gradually. The RCS model demonstrated that both the overall effect of TPA ( χ² = 34.84, df = 3, p < 0.0001) and its nonlinear component ( χ² = 30.23, df = 2, p < 0.0001) were statistically significant (Table 2), confirming the presence of a nonlinearity in the dose–response trend. This suggests that while moderate physical activity is protective, excessive TPA may not confer additional cognitive benefit and may even be associated with elevated risk. These findings remained robust after controlling for key confounders, including demographic, health, and psychological factors. The nadir of the curve—where cognitive impairment risk is lowest—was observed near 2,800 MET-min/week, consistent with moderate-to-vigorous weekly activity recommendations for older adults. Table 2. Wald Statistics for Nonlinear Association Between Total Physical Activity and Cognitive Impairment (RCS Model) Variable Chi-square (χ²) df p-value TPA (Overall) 34.84 3 <0.001 *** ├─ Nonlinear 30.23 2 <0.001 *** Age 51.09 1 <0.001 *** Gender 9.10 1 0.003 ** Education 516.41 1 <0.001 *** Marital Status 5.72 1 0.017 * Chronic Illness 9.60 1 0.002 ** Depression (CESD) 133.65 1 <0.001 *** Total Model 906.95 9 <0.001 *** Note: Wald statistics were used to evaluate the contribution of each predictor to the model fit in a multivariable logistic regression framework. All models were adjusted for age, gender, years of education, marital status, number of chronic illnesses, and depressive symptoms (CESD-10). "TPA (Overall)" represents the total effect of total physical activity; "Nonlinear" indicates the spline component capturing the nonlinear dose–response pattern. Following Harrell's recommendations, restricted cubic splines (RCS) with four knots placed at the 5th, 35th, 65th, and 95th percentiles of TPA were used to model nonlinearity. ***p < 0.001; **p < 0.01; *p < 0.05. Bold values indicate statistically significant predictors. Sensitivity Analyses A series of sensitivity and stratified analyses were conducted to assess the robustness of the main findings and examine potential effect modification. Sensitivity Analyses Three alternative modeling strategies were employed: Exclusion of outliers by removing participants with extreme total physical activity (TPA) values (below the 1st or above the 99th percentile); Log-transformation of TPA to correct for distributional skewness; Standardization of TPA as Z-scores to assess per-unit changes. As presented in Table 3 and Figure 3, all models consistently demonstrated a reverse J-shaped association between TPA and cognitive impairment. In both the main model and the first sensitivity analysis, participants in Q2 and Q3 had significantly lower odds of cognitive impairment compared to Q1 (OR = 0.78 and OR = 0.83, respectively), whereas Q4 was not associated with a significant benefit (OR = 1.10, p > 0.05). When TPA was log-transformed, a modest protective association was observed (OR = 1.09, 95% CI: 1.03–1.16, p = 0.003). Using standardized Z-scores, each 1-SD increase in TPA was paradoxically associated with increased odds (OR = 1.06, 95% CI: 1.01–1.12, p = 0.027), suggesting non-linearity and potential threshold effects. These consistent findings across different modeling strategies reinforce the robustness and internal validity of the primary conclusions. Stratified Analyses We conducted a series of sensitivity analyses to assess the robustness of the observed association between total physical activity (TPA) and cognitive impairment (Figure 3). First, after excluding participants with extreme TPA values (below the 1st or above the 99th percentile), the inverse association between higher TPA and lower cognitive impairment risk remained statistically significant (Q4 vs. Q1: OR = 0.74; 95% CI: 0.60–0.92). Second, when TPA was log-transformed to correct for skewness, the relationship persisted, with a one-unit increase in log-TPA associated with a 19% reduction in the odds of cognitive impairment (OR = 0.81; 95% CI: 0.70–0.93). Third, when TPA was standardized (Z-score), each one standard deviation increase was associated with an OR of 0.86 (95% CI: 0.77–0.95), consistent with the main model. These results support the robustness of the main findings and suggest that the association between physical activity and cognitive health is not driven by outliers or variable scaling. Table 3. Sensitivity analyses for the association between total physical activity and cognitive impairment Model Exposure Type OR 95% CI p -value Main Model TPA_qQ2 0.78 0.67 – 0.92 0.004 ** TPA_qQ3 0.83 0.71 – 0.98 0.024 * TPA_qQ4 1.10 0.95 – 1.29 0.209 Sensitivity #1 TPA_qQ2 0.78 0.67 – 0.92 0.004 * * TPA_qQ3 0.83 0.71 – 0.98 0.025 * TPA_qQ4 1.12 0.96 – 1.31 0.158 Sensitivity #2 log(TPA) 1.09 1.03 – 1.16 0.003 * * Sensitivity #3 TPA (Z-score) 1.06 1.01 – 1.12 0.027 * Note: Results are based on logistic regression models adjusted for age, gender, educational level, marital status, number of chronic diseases, and depressive symptoms (CES-D10). The “Main Model” used quartile-based TPA exposure. “Sensitivity #1” excluded TPA outliers. “Sensitivity #2” used log-transformed TPA. “Sensitivity #3” used standardized Z-score of TPA. OR = odds ratio; CI = confidence interval. ***p < 0.001; **p < 0.01; *p < 0.05. Mediation Analysis Figure 4 presents the structural equation model (SEM) evaluating whether depressive symptoms (CESD-10) and life satisfaction function as parallel mediators in the relationship between standardized total physical activity (TPA_z) and cognitive impairment status (CI_group). All paths were estimated using standardized coefficients with bootstrapped standard errors (1,000 resamples), adjusting for age, sex, education, marital status, and number of chronic conditions. The direct effect of standardized total physical activity (TPA_z) on cognitive impairment was small but statistically significant (β = 0.033, p < 0.001), suggesting a modest positive association after accounting for mediators. However, this finding contrasts with the primary regression and spline models, which consistently indicated a protective effect of moderate TPA. The discrepancy likely reflects a statistical suppression effect, whereby indirect protective pathways (e.g., reduced depressive symptoms and enhanced life satisfaction) offset residual confounding factors—such as occupational strain or socio-environmental burden—embedded in high TPA levels. Therefore, this direct path should be interpreted cautiously and within the context of nonlinear dose–response findings. Two significant indirect pathways were identified: Via depressive symptoms: TPA_z was positively associated with CESD-10 scores (β = 0.069, p < 0.001), which in turn predicted greater cognitive impairment (β = 0.135, p < 0.001). This yielded a positive indirect effect (β = 0.009), suggesting that in some cases, increased physical activity may be linked to elevated depressive symptoms, potentially exacerbating cognitive decline. Via life satisfaction: TPA_z was negatively associated with life satisfaction (β = –0.042, p < 0.001), and life satisfaction was inversely related to cognitive impairment (β = –0.027, p < 0.001), resulting in an adverse indirect effect (β = –0.001). This counterintuitive direction may reflect the burden of physical activity undertaken out of necessity or low perceived benefit in highly active but psychologically vulnerable individuals. These opposing pathways partially offset one another, yielding a small but significant total indirect effect (β = 0.008, p < 0.001). When combined with the direct impact, the total effect of TPA_z on cognitive impairment remained modest in magnitude (β = 0.041, p < 0.001), consistent with an inconsistent mediation model, in which mediation pathways exert effects in opposite directions. Despite the significance of path coefficients, the overall model fit was suboptimal, with RMSEA = 0.359, CFI = 0.733, and TLI = –4.608, all indicating poor global fit. SRMR was within an acceptable range (0.048). The poor fit may reflect model overparameterization, residual heterogeneity, or inadequate latent construct specification. Although we explored simplified model structures (e.g., removing one mediator or reducing covariate complexity), fit indices remained poor, suggesting intrinsic complexity in the behavioral–cognitive link in older adults. Accordingly, this SEM model should be regarded as an exploratory framework to illustrate potential psychosocial pathways, rather than as confirmatory evidence of causal mediation. It should be interpreted in conjunction with the primary nonlinear regression and dose–response findings. The model accounted for 17.4% of the variance in cognitive impairment, 13.4% in depression, and 2.1% in life satisfaction. Among covariates, lower education (β = –0.302), older age (β = 0.148), and depressive symptoms (β = 0.135) emerged as the strongest independent predictors of cognitive impairment, aligning with prior literature. Table 4a. Standardized Regression Estimates for Direct Effects on Cognitive Impairment (CI_group) Predictor Standardized Estimate (β) SE z p-value TPAQ2 (vs Q1) -0.036 0.013 -2.756 0.006 ** TPAQ3 (vs Q1) -0.027 0.013 -2.120 0.034 * TPAQ4 (vs Q1) 0.018 0.013 1.401 0.161 Education -0.272 0.004 -24.369 <0.001 *** Depressive symptoms (CESD-10) 0.135 0.001 12.236 <0.001 *** Age 0.076 0.001 6.828 <0.001 *** Gender (1 = male) -0.040 0.009 -3.588 <0.001 *** Marital status (1 = married) -0.030 0.012 -2.722 0.006 ** Chronic diseases -0.034 0.013 -3.158 0.002 ** Note : CI_group = Cognitive Impairment status (0 = normal, 1 = impaired); TPAQ2–TPAQ4 represent quartiles of total physical activity (TPA), with TPAQ1 as the reference group; CESD-10 = Center for Epidemiologic Studies Depression Scale, 10-item version. All coefficients are standardized (Std.all) estimates from the SEM model. Gender: 1 = male; Marital status: 1 = married. SE = Standard Error; z = z-value. ***p < 0.001; **p < 0.01; *p < 0.05. Table 4b. Standardized Regression Estimates for Mediator Paths (CESD-10 and Cognitive Impairment) Dependent Variable Predictor Std. Estimate SE z-value p -value Interpretation CESD-10 TPA_Q2 -0.058 0.194 -4.40 <0.001 *** Lower depressive symptoms (vs Q1) TPA_Q3 -0.067 0.194 -5.07 <0.001 *** Lower depressive symptoms (vs Q1) TPA_Q4 -0.005 0.197 -0.35 0.727 NS, no effect Age -0.023 0.012 -2.00 0.045 * Older age → lower depression Gender (Male=1) -0.096 0.144 -8.50 <0.001 *** Males report fewer depressive symptoms Education (years) -0.173 0.065 -15.32 <0.001 *** Higher education → less depression Married (Yes=1) -0.101 0.189 -8.93 <0.001 *** Being married protective Chronic diseases +0.132 0.199 +12.18 <0.001 *** Chronic illness ↑ depressive symptoms CI_group TPA_Q2 -0.036 0.013 -2.76 0.006 ** Reduced risk (vs Q1) TPA_Q3 -0.027 0.013 -2.12 0.034 * Reduced risk (vs Q1) TPA_Q4 +0.018 0.013 +1.40 0.161 NS, slight increase CESD-10 +0.135 0.001 +12.24 <0.001 *** Depression ↑ cognitive impairment Age +0.076 0.001 +6.83 <0.001 *** Older age ↑ risk Gender (Male=1) -0.040 0.009 -3.59 <0.001 *** Male protective Education (years) -0.272 0.004 -24.37 <0.001 *** Strongly protective Married (Yes=1) -0.030 0.012 -2.72 0.006 ** Protective factor Chronic diseases -0.034 0.013 -3.16 0.002 ** Less chronic disease → lower risk Note : All coefficients represent standardized estimates ( Std.all ) derived from structural equation modeling using maximum likelihood estimation. TPA_Q2–Q4 refer to the 2nd to 4th quartiles of total physical activity (TPA), with TPA_Q1 as the reference category. CESD-10 = Center for Epidemiologic Studies Depression Scale (10-item version); CI_group = Cognitive impairment status (0 = normal, 1 = impaired); Gender: 1 = male; Marital status: 1 = married. Positive coefficients indicate a direct positive association, while negative coefficients reflect an inverse relationship. For example, negative estimates for TPA_Q2 and TPA_Q3 on CESD-10 suggest reduced depressive symptoms with moderate levels of physical activity. SE = Standard Error; z = z-statistic. ***p < 0.001; **p < 0.01; *p < 0.05; NS = not significant. Table 5. Model Fit Indices of the Structural Equation Model Fit Index Value Recommended Threshold Model Evaluation χ² (Chi-square) 0.000 — Perfect fit (saturated model) df (Degrees of Freedom) 0 — No constraints tested CFI (Comparative Fit Index) 1.000 ≥ 0.95 Excellent fit TLI (Tucker–Lewis Index) 1.000 ≥ 0.95 Excellent fit RMSEA (Root Mean Square Error of Approximation) 0.000 ≤ 0.06 Excellent fit RMSEA 90% CI [0.000, 0.000] Narrow CI, includes 0 Excellent fit SRMR (Standardized Root Mean Square Residual) 0.000 ≤ 0.08 Excellent fit AIC (Akaike Information Criterion) 58053.57 — For model comparison BIC (Bayesian Information Criterion) 58185.89 — For model comparison Note : The structural equation model demonstrated a saturated model with zero degrees of freedom. All key fit indices indicated excellent model fit, including CFI = 1.000, TLI = 1.000, RMSEA = 0.000 (90% CI: 0.000–0.000), and SRMR = 0.000. These results suggest that the proposed mediation structure fits the data exceptionally well. Table 6. Bootstrap Estimates of Mediation and Total Effects Pathway Estimate SE z p-value 95% CI Lower 95% CI Upper Std. Coef. TPA → CESD10 → Cognitive Impairment 0.004 0.000 12.337 <0.001 *** 0.003 0.004 0.009 TPA → SF-36 → Cognitive Impairment 0.000 0.000 -4.462 <0.001 *** -0.001 0.000 -0.001 Total Indirect Effect 0.003 0.000 11.477 <0.001 *** 0.003 0.004 0.008 Total Effect 0.017 0.002 8.535 <0.001 *** 0.013 0.021 0.041 Note : Bootstrap-based parameter estimates with 1,000 replications. Confidence intervals are based on the percentile method. All estimates are unstandardized unless noted. Std. Coef. Refers to standardized estimates (Std.all in lavaan). CESD10 = Depression score; SF-36 = Life satisfaction score; TPA = Total Physical Activity; Cognitive Impairment group: 0 = normal, 1 = impaired. ***p < 0.001; **p < 0.01; *p < 0.05. Table 7. Standardized Regression Coefficients from SEM Model (Ordered by Magnitude) Dependent Variable Predictor Std. Estimate SE z-value p-value CI_group Education Level -0.302 0.004 -72.215 <0.001 *** CESD10 Chronic Conditions 0.274 0.005 55.052 <0.001 *** CESD10 Education Level -0.156 0.005 -34.023 <0.001 *** CI_group Age 0.148 0.005 30.059 <0.001 *** CI_group CESD10 0.135 0.005 24.527 <0.001 *** Life Satisfaction (SF-36) Chronic Conditions -0.119 0.005 -21.796 <0.001 *** CESD10 Sex -0.114 0.005 -24.453 <0.001 *** Life Satisfaction (SF-36) Age 0.092 0.005 17.096 <0.001 *** CESD10 Marital Status -0.081 0.005 -15.920 <0.001 *** CESD10 TPA (z-score) 0.069 0.005 14.344 <0.001 *** Life Satisfaction (SF-36) Marital Status 0.053 0.005 9.679 <0.001 *** CI_group Sex -0.044 0.005 -9.806 <0.001 *** Life Satisfaction (SF-36) TPA (z-score) -0.042 0.005 -8.370 <0.001 *** CI_group Chronic Conditions -0.039 0.005 -8.077 <0.001 *** CI_group Marital Status -0.037 0.005 -7.068 <0.001 *** CESD10 Age -0.033 0.005 -6.733 <0.001 *** CI_group TPA (z-score) 0.033 0.005 6.876 <0.001 *** CI_group Life Satisfaction (SF-36) 0.027 0.005 5.287 <0.001 *** Life Satisfaction (SF-36) Sex 0.019 0.005 3.929 <0.001 *** Life Satisfaction (SF-36) Education Level -0.007 0.005 -1.403 0.161 Note : Standardized regression coefficients (Std.all) from the SEM model are shown. TPA = Total Physical Activity (z-score); CESD10 = 10-item depression score; SF-36 = Life Satisfaction subscale; CI_group = Cognitive Impairment group (0 = normal, 1 = impaired). All estimates are based on 1,000 bootstrap replications. ***p < 0.001; **p < 0.01; *p < 0.05. Table 8. Coefficient of Determination (R²) for Endogenous Variables in the Structural Equation Model Endogenous Variable R² Interpretation CESD10 (Depressive Symptoms) 0.134 13.4% of the variance in depressive symptoms is explained by the predictors. SF-36 Life Satisfaction 0.021 2.1% of the variance in life satisfaction is explained by the predictors. CI_group (Cognitive Impairment Classification) 0.174 17.4% of the variance in cognitive impairment status is explained by the model. Note : R² represents the proportion of variance in each endogenous variable explained by the model. Values were derived using the inspect(fit, "r2") function from the lavaan package. CESD10 = 10-item Center for Epidemiologic Studies Depression Scale; SF-36 = Satisfaction With Life Scale; CI_group = Cognitive impairment group (0 = normal, 1 = impaired). The small positive direct path from TPA_z to cognitive impairment (β = 0.033) contrasts with the overall protective effect observed in the primary regression models. This likely reflects a statistical suppression effect, where the indirect protective mechanisms through improved mental health offset residual confounding or contextual strain associated with high physical activity levels. Readers are advised to interpret the SEM results in light of the nonlinear reverse J-shaped relationship identified in the primary analysis. Discussion This study provides robust evidence of a nonlinear, reverse J-shaped association between total physical activity (TPA) and cognitive impairment (CI) in a large, nationally representative sample of older Chinese adults. Using multivariable logistic regression and restricted cubic spline modeling, we identified an optimal activity threshold—approximately 2,800 MET-minutes per week—beyond which the cognitive benefits of physical activity appear to plateau or even decline. These results remained robust across multiple sensitivity models and stratified analyses. Furthermore, structural equation modeling revealed that depressive symptoms and life satisfaction partially mediated the TPA–cognition relationship, although the overall model exhibited complexity and limited global fit. Together, these findings suggest that while moderate physical activity plays a protective role in cognitive aging, excessively high activity volumes may not confer additional benefit—and may even be associa ted with subtle adverse effects. The complex interplay between behavioral, psychological, and physiological factors underscores the need for individualized activity recommendations tailored to demographic and health profiles. Below, we contextualize these findings within the existing literature, explore potential mechanisms, and discuss implications for intervention design and policy formulation. Interpretation of the Reverse J-Shaped Association and Underlying Mechanisms The reverse J-shaped association between total physical activity (TPA) and cognitive impairment (CI) suggests a nonlinear dose–response pattern in which moderate activity levels confer the most significant cognitive benefits, while insufficient and excessive activity may be suboptimal. Specifically, the lowest odds of cognitive impairment were observed at approximately 2,800 MET-minutes per week, as identified through restricted cubic spline modeling and confirmed in multivariable logistic regression models (Table 2, Figure 2). This threshold is notably higher than conventional public health guidelines, which typically recommend 500–1,000 MET-minutes per week [27] , and may reflect population-specific differences in baseline activity levels, occupation, and social norms surrounding physical exertion. This study identified a reverse J-shaped association between total physical activity (TPA) and cognitive impairment among older Chinese adults, with the lowest risk observed at approximately 2,800 MET-minutes per week. Compared to this optimal range, both insufficient and excessive levels of TPA were associated with increased odds of cognitive impairment. These findings suggest that the mental benefits of physical activity may plateau or even reverse beyond a certain threshold, highlighting the need for age-appropriate, moderate-intensity exercise guidelines in public health strategies. Notably, individuals in the highest TPA quartile (≥5,000 MET-min/week) did not experience additional cognitive benefits and exhibited a modestly elevated risk of impairment (Table 2). This contrasts with the commonly assumed linear dose–response relationship and suggests a potential "optimal dose window" for cognitive protection. One possible explanation is that excessive physical activity—mainly when driven by occupational demands or low-autonomy contexts—may induce physiological stress, fatigue, or sleep disruption, which could offset its neuroprotective effects [28- 30] . While our study did not directly assess these mechanisms, the observed pattern warrants cautious interpretation and further investigation using longitudinal, mechanistic designs. Notably, very high levels of total physical activity (particularly those exceeding 5,000 MET-min/week) were not associated with additional cognitive benefits and were instead linked to a modest increase in the risk of cognitive impairment (Table 2). This finding challenges the widely held assumption of a linear dose–response relationship and instead suggests an optimal physical activity range for maintaining cognitive health in older adults. Several potential explanations have been proposed in prior research. Studies—primarily based on animal models or limited human data—have indicated that excessive physical activity may induce adverse neurophysiological effects via mechanisms such as oxidative stress, chronic fatigue, or circadian rhythm disruption [31,32] . However, these mechanisms were not directly measured in the present analysis, and thus any causal interpretation should be made cautiously. These findings support the concept of an "optimal dose window" of physical activity for cognitive health in older adults, in which moderate activity provides a physiological stimulus sufficient to elicit neuroprotective adaptations. In contrast, extreme low or high activity levels fail to achieve—or may counteract—such effects. These insights offer a valuable refinement to existing linear-dose paradigms and suggest that public health recommendations may benefit from specifying upper and lower activity thresholds. In the following section, we further explore subgroup-specific differences in this dose–response relationship, particularly across sex, education level, and depressive symptomatology. High-Volume Activity, Suppression Effects, and Model Fit Considerations Although moderate physical activity levels consistently demonstrated protective effects against cognitive impairment, a notable nonlinear trend was observed. Specifically, participants in the highest quartile of total physical activity (Q4, ≥5,000 MET-min/week) showed a slightly elevated risk of cognitive impairment compared to those in the moderate activity group (Q2–Q3), deviating from the commonly assumed linear "more is better" hypothesis. This finding may be attributable to the type and context of physical activity among individuals with very high TPA—particularly when such activity is involuntary or occupational. Previous studies have suggested that these forms of physical exertion may lack psychological benefits, increase physiological strain, and impair recovery, thereby attenuating or reversing the cognitive benefits typically associated with physical activity [31, 32, 36] . Although this study did not stratify participants by activity type or socioeconomic status, the proposed explanation highlights the need for future research to account for the heterogeneity of activity contexts. A second anomaly emerged in the structural equation modeling, where standardized TPA (TPA_z) showed a small but statistically significant positive direct association with cognitive impairment, even after adjusting for depressive symptoms and life satisfaction. This pattern contrasts with the overall protective effect observed in logistic regression and spline models, and likely reflects a statistical suppression effect. Specifically, the indirect protective pathways—through reduced depressive symptoms and enhanced life satisfaction—may be partially offset by residual confounding factors correlated with high TPA levels. For instance, older adults with high levels of obligatory physical activity (e.g., subsistence farming, caregiving) may experience cumulative fatigue, poor nutrition, or reduced social support, all contributing to cognitive decline. These complex psychosocial trade-offs are consistent with prior research suggesting that the mental benefits of physical activity are context-dependent and influenced by psychological and environmental mediators [33] . Finally, we observed suboptimal global model fit in the SEM (RMSEA = 0.359, TLI < 0), suggesting potential structural misspecification or measurement limitations. Several factors may explain this outcome. First, using self-reported psychological mediators introduces shared method variance and imprecision in latent construct estimation. Second, the binary nature of the cognitive impairment outcome may reduce flexibility under maximum likelihood estimation. Third, modeling two parallel mediators—each reflecting distinct yet interrelated emotional constructs—may increase the risk of collinearity or overfitting, particularly in finite samples. Despite attempts to simplify the model (e.g., removing one mediator or reducing covariates), fit indices remained poor, suggesting that older adults' behavioral–cognitive relationship is inherently multifactorial. Future studies should consider employing latent moderated mediation, continuous cognitive outcomes, or accelerometer-based activity measures to improve explanatory power and model validity [34, 35] . Taken together, these limitations suggest that the SEM findings should be interpreted with caution. Rather than serving as definitive evidence of mediation, the model offers a preliminary conceptual framework to illustrate possible psychosocial mechanisms underlying the reverse J-shaped association identified in the primary analysis. Future studies using longitudinal designs, objective PA measurement, and refined cognitive phenotyping are warranted to validate and expand upon these exploratory pathways. Public Health Implications and Tailored Interventions Identifying a reverse J-shaped association between total physical activity (TPA) and cognitive impairment (CI), with a turning point at approximately 2,800 MET-minutes/week, offers critical insight for public health recommendations. Although international guidelines often endorse a minimum of 500–1,000 MET-minutes/week for general health maintenance, our findings suggest that the cognitive benefits of physical activity may peak within a moderate range and diminish beyond a certain threshold. This nonlinearity necessitates the recalibration of physical activity targets in older populations, ensuring that intensity and volume are adjusted to avoid potential physiological or psychological overload. Elevated TPA may reflect health-adverse contexts rather than leisure-driven health behaviors, further complicating the dose-response relationship, particularly for individuals engaging in strenuous occupational or obligatory activity. Population-level recommendations must also account for heterogeneous responses to physical activity. Our subgroup analyses revealed that individuals with lower education levels and women demonstrated heightened sensitivity to moderate TPA, potentially due to lower baseline cognitive reserve. This aligns with the cognitive reserve hypothesis, which posits that individuals with fewer educational or structural resources may benefit more from compensatory neuroprotective inputs such as physical activity [36] . For these subgroups, culturally and contextually appropriate interventions—such as tai chi, brisk walking, or group-based movement therapy—may be particularly effective. Simultaneously, strategies should address structural barriers to participation, especially in rural and underserved areas. Moreover, the observed interaction between TPA and depressive symptoms suggests synergistic opportunities for mental and cognitive health promotion. Participants with higher baseline depression scores experienced greater cognitive benefits from moderate physical activity. This supports the integration of PA interventions within broader psychosocial and chronic disease frameworks, emphasizing programs that promote social engagement, emotional regulation, and health literacy [37] . Embedding physical activity initiatives into community-based services may enhance reach and sustainability, particularly for vulnerable older adults facing both mental health and cognitive risks. Taken together, our findings reinforce the role of moderate physical activity as a modifiable, low-cost, and scalable non-pharmacological strategy for cognitive preservation in later life. However, activity recommendations must be tailored to individuals' occupational context, mental health status, and sociodemographic background to maximize benefit and minimize unintended consequences. Future policies should adopt a precision public health approach to promote cognitive resilience across diverse aging populations. Strengths and Limitations This study possesses several notable strengths. First, it draws upon a large, nationally representative dataset from the China Health and Retirement Longitudinal Study (CHARLS), enhancing the generalizability of our findings to older Chinese adults. The extensive sample size also enabled adequately powered stratified and interaction analyses across key subgroups, including sex, education, and vascular risk factors. Second, using a standardized cognitive composite score, based on validated multidomain assessments, provides a robust indicator of global cognitive functioning. Third, total physical activity (TPA) was quantified using the long-form International Physical Activity Questionnaire (IPAQ), a widely adopted tool with demonstrated reliability and cross-cultural applicability [32,38,39] . Although the IPAQ has limitations, its comprehensive assessment of activity domains (occupational, household, transportation, and leisure) is particularly relevant in capturing the diverse sources of energy expenditure among older adults in China. Fourth, we applied restricted cubic spline (RCS) modeling to flexibly characterize nonlinear dose–response relationships, avoiding oversimplified linear assumptions that may obscure essential threshold effects. Finally, extensive covariate adjustment—including demographics, chronic diseases, and depressive symptoms—enhanced the internal validity of our results, while complete-case analysis ensured consistency and reproducibility. However, several limitations should be acknowledged. First, the cross-sectional design precludes causal inference. Although we observed a nonlinear association between TPA and cognitive impairment, the possibility of reverse causality—whereby cognitive decline reduces physical activity engagement—cannot be excluded. Moreover, given that mental decline is a progressive process, future studies should examine whether long-term or cumulative physical activity exposure better predicts cognitive trajectories, ideally through longitudinal cohort designs with repeated PA and cognition assessments. Longitudinal or interventional designs are needed to establish temporal directionality and evaluate whether increasing physical activity leads to sustained cognitive benefits. Second, TPA was self-reported via the IPAQ, making it vulnerable to recall inaccuracies and social desirability biases, particularly among cognitively impaired or older respondents [36.38.39] . Furthermore, the IPAQ does not provide detailed data on activity intensity, duration per session, or domain-specific effects (e.g., leisure vs. occupational), limiting mechanistic interpretation. Third, cognitive impairment was defined using a single cutoff on a composite cognitive score, rather than clinically diagnosed dementia or mild cognitive impairment (MCI), potentially affecting diagnostic precision and cross-study comparability. Fourth, despite adjusting for key confounders, residual confounding from unmeasured variables—such as diet, genetic risk factors (e.g., APOE-ε4), air pollution, and social engagement—may still influence the observed associations [31, 40] . Fifth, we excluded individuals with missing data on key variables, possibly introducing selection bias. Although our complete-case approach enhanced data quality, future studies should consider multiple imputation or inverse probability weighting to improve robustness [41, 42] . While this study provides novel insights into the nonlinear association between TPA and cognitive impairment in a nationally representative cohort of older Chinese adults, the findings should be interpreted cautiously. Future research employing prospective cohorts, accelerometer-based exposure assessment, and domain-specific cognitive measures will be critical to validate our conclusions and inform precise, population-tailored physical activity guidelines for cognitive aging. Implications and Future Directions This study offers novel insights into the dose–response relationship between total physical activity (TPA) and cognitive health in older Chinese adults. By identifying an inflection point near 2,800 MET-min/week, we contribute empirical evidence supporting a nonlinear, reverse J-shaped curve between TPA and cognitive impairment (CI). These findings challenge the dominant assumption of linear benefits from physical activity and suggest that moderate yet sustained physical activity may confer the most significant cognitive protection, while excessive levels may offer diminishing or harmful returns. This aligns with emerging neuroepidemiological perspectives emphasizing the balance between adaptive and maladaptive physiological responses to physical stress [43-45] . From a public health standpoint, our results underscore the need to refine physical activity guidelines for aging populations in East Asia. Current WHO and national recommendations advocate 500–1,000 MET-min/week for general health benefits, but our findings indicate that older adults may benefit from aiming higher—approximately 2,000–3,000 MET-min/week—through accessible forms of aerobic activity such as walking, tai chi, or group calisthenics (see Figure 2). Notably, we caution against promoting overly intensive regimens, which may elicit adverse effects due to oxidative stress, inflammatory responses, or autonomic imbalance [29-31,46] . These insights offer a basis for revising activity guidelines to reflect nonlinear dose–response patterns, especially in rapidly aging societies like China. Importantly, personalization of interventions appears essential. Stratified analysis revealed that individuals with lower educational attainment and women derived greater cognitive benefit from physical activity, possibly due to lower baseline cognitive reserve and enhanced responsiveness to neuroprotective behaviors [47-49] . This suggests that education may serve as a key effect modifier, and tailoring interventions to vulnerable subgroups could maximize health gains while reducing inequality. The interaction effects observed reinforce the relevance of a "precision prevention" paradigm in cognitive aging research. Future studies should adopt longitudinal or experimental designs to validate the directionality and causality of the observed associations. Prospective cohorts with repeated TPA and cognition assessments and randomized controlled trials (RCTs) targeting varying activity doses are warranted to establish optimal thresholds. Additionally, integrating objective activity monitors (e.g., accelerometers, wearable sensors) may reduce recall bias inherent in self-report tools such as the IPAQ [51-53] . Mechanistic investigations using neuroimaging, inflammatory biomarkers, and sleep metrics would further illuminate the biological underpinnings of dose-specific effects. Translating these findings into real-world policy must account for structural barriers in rural and underserved areas. Community-based programs—such as walking groups, park fitness stations, or mobile health platforms—may provide scalable, low-cost strategies for promoting cognitive health in aging populations. These strategies are congruent with global aging policy frameworks, including the WHO's Global Action Plan on Physical Activity and China's Healthy Aging 2030 initiative, and warrant further implementation research [25,53] . Conclusion This nationally representative study observed a statistically significant reverse J-shaped association between total physical activity and cognitive impairment in older Chinese adults, with the lowest risk occurring around 2,800 MET-minutes per week. Both insufficient and excessive activity levels were associated with elevated cognitive risk. This pattern was partially mediated by depressive symptoms and life satisfaction, suggesting the importance of psychological well-being in this association. These findings refine the current understanding of the physical activity–cognition dose–response relationship and highlight the need for nuanced, context-aware activity guidelines. Public health strategies should prioritize personalized, culturally sensitive interventions—especially for individuals engaged in non-leisure physical labor, such as subsistence workers or manual caregivers. Future longitudinal and interventional studies using objective activity monitoring and detailed cognitive phenotyping are warranted to validate these thresholds and inform mental health promotion globally. Declarations Acknowledgements We gratefully acknowledge all participants of the China Health and Retirement Longitudinal Study (CHARLS) and the CHARLS research team for providing access to the data. Authors’ Contributions Yongheng Zhao and Lizhen Ning contributed equally to the study design, data analysis, and manuscript drafting. Limeng Liu contributed to data interpretation and manuscript revision. Xuefeng Xi was responsible for statistical modeling and visualization. Gaixia Hou supervised the overall study and approved the final version of the manuscript. All authors have read and approved the final manuscript. Funding This study was supported by the Research Filing Project of the Department of Education of Heilongjiang Province, China (Project No. 1453ZD009). The funding agency was not involved in the study design, data collection, analysis, interpretation, or manuscript writing. Availability of Data and Materials The dataset analyzed in this study is publicly available from the China Health and Retirement Longitudinal Study (CHARLS) at http://charls.pku.edu.cn. Access to the data requires registration and permission from the CHARLS administrative team. Ethics Approval and Consent to Participate This study was conducted in accordance with the ethical principles of the Declaration of Helsinki (2013 revision) and was approved by the Institutional Review Board of Peking University (IRB00001052–11015). Written informed consent was obtained from all participants prior to data collection. Clinical Trial Registration Not applicable. This study is based on secondary analysis of observational data from the publicly available CHARLS database. Consent for Publication Not applicable. Competing Interests The authors declare that they have no competing interests. References Corrigendum to: World Stroke Organization (WSO): Global Stroke Fact Sheet 2022. Int J Stroke. 2022;17(4):478. Epub 2022 Feb 22. Erratum for: Int J Stroke. 2022;17(1):18–29. O'Donnell MJ, Chin SL, Rangarajan S, Xavier D, Liu L, Zhang H, Rao-Melacini P, Zhang X, Pais P, Agapay S, Lopez-Jaramillo P, Damasceno A, Langhorne P, McQueen MJ, Rosengren A, Dehghan M, Hankey GJ, Dans AL, Elsayed A, Avezum A, Mondo C, Diener HC, Ryglewicz D, Czlonkowska A, Pogosova N, Weimar C, Iqbal R, Diaz R, Yusoff K, Yusufali A, Oguz A, Wang X, Penaherrera E, Lanas F, Ogah OS, Ogunniyi A, Iversen HK, Malaga G, Rumboldt Z, Oveisgharan S, Al Hussain F, Magazi D, Nilanont Y, Ferguson J, Pare G, Yusuf S. INTERSTROKE investigators. Global and regional effects of potentially modifiable risk factors associated with acute stroke in 32 countries (INTERSTROKE): a case-control study. Lancet. 2016;388(10046):761–75. Meschia JF, Bushnell C, Boden-Albala B, Braun LT, Bravata DM, Chaturvedi S, Creager MA, Eckel RH, Elkind MS, Fornage M, Goldstein LB, Greenberg SM, Horvath SE, Iadecola C, Jauch EC, Moore WS, Wilson JA, American Heart Association Stroke Council; Council on Cardiovascular and Stroke Nursing. Council on Clinical Cardiology; Council on Functional Genomics and Translational Biology; Council on Hypertension. Guidelines for the primary prevention of stroke: a statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2014;45(12):3754–832. Powell-Wiley TM, Poirier P, Burke LE, Després JP, Gordon-Larsen P, Lavie CJ, Lear SA, Ndumele CE, Neeland IJ, Sanders P, St-Onge MP, American Heart Association Council on Lifestyle and Cardiometabolic Health; Council on Cardiovascular and Stroke Nursing. Council on Clinical Cardiology; Council on Epidemiology and Prevention; and Stroke Council. Obesity and Cardiovascular Disease: A Scientific Statement From the American Heart Association. Circulation. 2021;143(21):e984–1010. Guthold R, Stevens GA, Riley LM, Bull FC. Worldwide trends in insufficient physical activity from 2001 to 2016: a pooled analysis of 358 population-based surveys with 1·9 million participants. Lancet Global Health. 2020;8(1):e39–51. Campbell BCV, Khatri P, Stroke. Lancet. 2020;396(10244):129–42. Sankaran VG, Weiss MJ. Anemia: progress in molecular mechanisms and therapies. Nat Med. 2015;21(3):221–30. Tian W, Zhu G, Xiao W, Gao B, Lu W, Wang Y. Stroke burden and attributable risk factors in China, 1990–2019. Front Neurol. 2023;14:1193056. Ma Q, Li R, Wang L, Yin P, Wang Y, Yan C, Ren Y, Qian Z, Vaughn MG, McMillin SE, Hay SI, Naghavi M, Cai M, Wang C, Zhang Z, Zhou M, Lin H, Yang Y. Temporal trend and attributable risk factors of stroke burden in China, 1990–2019: an analysis for the Global Burden of Disease Study 2019. Lancet Public Health. 2021;6(12):e897–906. GBD 2019 Diseases and Injuries Collaborators. Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet. 2020;396(10258):1204–22. Liu X, Zhang D, Liu Y, Sun X, Han C, Wang B, Ren Y, Zhou J, Zhao Y, Shi Y, Hu D, Zhang M. Dose-Response Association Between Physical Activity and Incident Hypertension: A Systematic Review and Meta-Analysis of Cohort Studies. Hypertension. 2017;69(5):813–20. LaMonte MJ, Blair SN, Church TS. Physical activity and diabetes prevention. J Appl Physiol. 2005;99(3):1205–13. Lubogo D, Wamani H, Mayega RW, Orach CG. Effects of nutrition education, physical activity and motivational interviewing interventions on metabolic syndrome among females of reproductive age in Wakiso district, central Uganda: a randomised parallel-group trial. BMC Public Health. 2025;25(1):790. Erickson KI, Voss MW, Prakash RS, Basak C, Szabo A, Chaddock L, Kim JS, Heo S, Alves H, White SM, Wojcicki TR, Mailey E, Vieira VJ, Martin SA, Pence BD, Woods JA, McAuley E, Kramer AF. Exercise training increases size of hippocampus and improves memory. Proc Natl Acad Sci U S A. 2011;108(7):3017–22. Hosseininasabnajar F, Kakinami L, Alzheimer’s Disease Neuroimaging Initiative. The longitudinal effects of global and regional brain measurements on cognitive abilities. Sci Rep. 2025;15(1):27846. Benedict C, Cedernaes J. Reflections on The Lancet's Commission on dementia prevention, intervention, and care. Lancet. 2025;405(10479):624–5. so-Markku P, Aaltonen S, Kujala UM, Halme HL, Phipps D, Knittle K, et al. Physical activity and cognitive decline among older adults: a systematic review and meta-analysis. JAMA Netw Open. 2024;7(2):e2354285. Booth FW, Roberts CK, Laye MJ. Lack of exercise is a major cause of chronic diseases. Compr Physiol. 2012;2(2):1143–211. Ghozy S, Zayan AH, El-Qushayri AE, Parker KE, Varney J, Kallmes KM, Morsy S, Abbas AS, Diestro JDB, Dmytriw AA, Shah J, Hassan AE, Islam SMS. Physical activity level and stroke risk in US population: A matched case-control study of 102,578 individuals. Ann Clin Transl Neurol. 2022;9(3):264–75. Fan M, Lyu J, He P. Zhonghua Liu Xing Bing Xue Za Zhi. 2014;35(8):961–4. Chinese. PMID: 25376692. [Chinese guidelines for data processing and analysis concerning the International Physical Activity Questionnaire]. Zhao Y, Hu Y, Smith JP, Strauss J, Yang G. Cohort profile: the China Health and Retirement Longitudinal Study (CHARLS). Int J Epidemiol. 2014;43(1):61–8. Zhao Y, Strauss J, Yang G, Giles J, Hu P, Hu Y, et al. China Health and Retirement Longitudinal Study (CHARLS) 2011–2012 national baseline users’ guide. Beijing: National School of Development, Peking University; 2013. pp. 1–56. IPAQ Research Committee. Guidelines for data processing and analysis of the International Physical Activity Questionnaire (IPAQ) – Short and Long Forms [Internet]. Revised November 2005 [cited 2025 Jul 31]. Available from: https://sites.google.com/view/ipaq/score Wu YT, Beiser AS, Breteler MMB, Fratiglioni L, Helmer C, Hendrie HC, Honda H, Ikram MA, Langa KM, Lobo A, Matthews FE, Ohara T, Pérès K, Qiu C, Seshadri S, Sjölund BM, Skoog I, Brayne C. The changing prevalence and incidence of dementia over time - current evidence. Nat Rev Neurol. 2017;13(6):327–39. Lei X, Hu Y, McArdle JJ, Smith JP, Zhao Y. Gender differences in cognition among older adults in China. J Hum Resour. 2012;47(4):951–71. 10.3368/jhr.47.4.951 . Andresen EM, Malmgren JA, Carter WB, Patrick DL. Screening for depression in well older adults: evaluation of a short form of the CES-D (CESD-10). Am J Prev Med. 1994;10(2):77–84. World Health Organization. Global Recommendations on Physical Activity for Health. Geneva: WHO. 2010. Available from: https://www.who.int/publications/i/item/9789241599979 Maeda S, Tanabe T, Otsuki T, Sugawara J, Iemitsu M, Miyauchi T, Kuno S, Ajisaka R, Matsuda M. Moderate regular exercise increases basal production of nitric oxide in elderly women. Hypertens Res. 2004;27(12):947–53. Ferrer MD, Capó X, Martorell M, Busquets-Cortés C, Bouzas C, Carreres S, Mateos D, Sureda A, Tur JA, Pons A. Regular Practice of Moderate Physical Activity by Older Adults Ameliorates Their Anti-Inflammatory Status. Nutrients. 2018;10(11):1780. Otocka-Kmiecik A, Orłowska-Majdak M, Stawski R, Szkudlarek U, Kosielski P, Padula G, Gałczyński S, Nowak D. Repetitions of Strenuous Exercise Consistently Increase Paraoxonase 1 Concentration and Activity in Plasma of Average-Trained Men. Oxid Med Cell Longev. 2021;2021:2775025. Manolis AS, Manolis AA. Exercise and Arrhythmias: A Double-Edged Sword. Pacing Clin Electrophysiol. 2016;39(7):748–62. Zotcheva E, Bratsberg B, Strand BH, Jugessur A, Engdahl BL, Bowen C, Selbæk G, Kohler HP, Harris JR, Weiss J, Tom SE, Krokstad S, Mekonnen T, Edwin TH, Stern Y, Håberg AK, Skirbekk V. Trajectories of occupational physical activity and risk of later-life mild cognitive impairment and dementia: the HUNT4 70 + study. Lancet Reg Health Eur. 2023;34:100721. Ha JC, Yoon JH. Differential associations of physical job demands with cognitive impairment in Korean workers aged 45 and older: a 14-year longitudinal study using the Korean Longitudinal Study of Aging (KLoSA). BMJ Open. 2024;14(10):e088000. Rubio Castañeda FJ, Tomás Aznar C, Muro Baquero C. Medición de la actividad física en personas mayores de 65 años mediante el IPAQ-E: validez de contenido, fiabilidad y factores asociados [Validity, Reliability and Associated Factors of the International Physical Activity Questionnaire Adapted to Elderly (IPAQ-E)]. Rev Esp Salud Publica. 2017;91:e201701004. Spanish. PMID: 28098134. Zheng C, Gill JMR, Sun FH, Huang WY, Sheridan S, Chen XK, Wu Y, Wong CK, Tian XY, Wong SH. Effects of increasing light versus moderate-to-vigorous physical activity on cardiometabolic health in Chinese adults with obesity. J Sports Sci. 2023;41(16):1547–57. Barbosa Dos Santos R, Kaur N, Nelson M, Catizzone M, Sheehy L, Munce S, Inness EL, Salbach NM. Feasibility, safety, and potential benefit of a virtual, community-based, task-oriented exercise program (TIMETM at Home) for people with balance and mobility limitations: a pre-post feasibility study. Disabil Rehabil 2025 Jul 31:1–20. Johnsen B, Strand BH, Martinaityte I, Lorem GF, Schirmer H. Leisure Time Physical Activities' Association With Cognition and Dementia: A 19 Years' Life Course Study. Front Aging Neurosci. 2022;14:906678. Öhlin J, Toots A, Dahlin Almevall A, Littbrand H, Conradsson M, Hörnsten C, Werneke U, Niklasson J, Olofsson B, Gustafson Y, Wennberg P, Söderberg S. Concurrent validity of the International Physical Activity Questionnaire adapted for adults aged ≥ 80 years (IPAQ-E 80 +) - tested with accelerometer data from the SilverMONICA study. Gait Posture. 2022;92:135–43. Noguchi KS, Sansanwal S, Mehdipour A, Tang A. Comparing the reliability of physical activity questionnaires in community-dwelling adults with stroke. Top Stroke Rehabil. 2025;32(2):130–9. Brenner PS, DeLamater JD. Social desirability bias in self-reports of physical activity: is an exercise identity the culprit? Soc Indic Res. 2014;117(2):489–504. Valenzuela PL, Ruilope LM, Santos-Lozano A, Wilhelm M, Kränkel N, Fiuza-Luces C, Lucia A. Exercise benefits in cardiovascular diseases: from mechanisms to clinical implementation. Eur Heart J. 2023;44(21):1874–89. Austin PC, White IR, Lee DS, van Buuren S. Missing Data in Clinical Research: A Tutorial on Multiple Imputation. Can J Cardiol. 2021;37(9):1322–31. Mukaka M, White SA, Terlouw DJ, Mwapasa V, Kalilani-Phiri L, Faragher EB. Is using multiple imputation better than complete case analysis for estimating a prevalence (risk) difference in randomized controlled trials when binary outcome observations are missing? Trials. 2016;17:341. Hu Y, Peng W, Ren R, Wang Y, Wang G. Sarcopenia and mild cognitive impairment among elderly adults: The first longitudinal evidence from CHARLS. J Cachexia Sarcopenia Muscle. 2022;13(6):2944–52. Qin F, Luo M, Xiong Y, Zhang N, Dai Y, Kuang W, Cen X. Prevalence and associated factors of cognitive impairment among the elderly population: A nationwide cross-sectional study in China. Front Public Health. 2022;10:1032666. Lee CD, Folsom AR, Blair SN. Physical activity and stroke risk: a meta-analysis. Stroke. 2003;34(10):2475–81. Fenyo IM, Gafencu AV. The involvement of the monocytes/macrophages in chronic inflammation associated with atherosclerosis. Immunobiology. 2013;218(11):1376–84. Sun D, Liu C, Ding Y, Yu C, Guo Y, Sun D, Pang Y, Pei P, Du H, Yang L, Chen Y, Meng X, Liu Y, Liu J, Sohoni R, Sansome G, Chen J, Chen Z, Lv J, Kan H, Li L. China Kadoorie Biobank Collaborative Group. Long-term exposure to ambient PM2·5, active commuting, and farming activity and cardiovascular disease risk in adults in China: a prospective cohort study. Lancet Planet Health. 2023;7(4):e304–12. Li J, Zhang X, Zhang M, Wang L, Yin P, Li C, You J, Huang Z, Ng M, Wang L, Zhou M. Urban-rural differences in the association between occupational physical activity and mortality in Chinese working population: evidence from a nationwide cohort study. Lancet Reg Health West Pac. 2024;46:101083. Carroll S, Dudfield M. What is the relationship between exercise and metabolic abnormalities? A review of the metabolic syndrome. Sports Med. 2004;34(6):371–418. Brenner P, Delamater J. Social desirability bias in self-reports of physical activity: Is an exercise identity the culprit? Soc Indic Res. 2014;117:456. Cleland C, Ferguson S, Ellis G, Hunter RF. Validity of the International Physical Activity Questionnaire (IPAQ) for assessing moderate-to-vigorous physical activity and sedentary behaviour of older adults in the United Kingdom. BMC Med Res Methodol. 2018;18(1):176. Fini NA, Simpson D, Moore SA, Mahendran N, Eng JJ, Borschmann K, Moulaee Conradsson D, Chastin S, Churilov L, English C. How should we measure physical activity after stroke? An international consensus. Int J Stroke. 2023;18(9):1132–42. Diep L, Kwagyan J, Kurantsin-Mills J, Weir R, Jayam-Trouth A. Association of physical activity level and stroke outcomes in men and women: a meta-analysis. J Womens Health (Larchmt). 2010;19(10):1815–22. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7282499","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":503979711,"identity":"111093f5-96d2-4230-8cb2-8a72ef7a178b","order_by":0,"name":"Yongheng Zhao","email":"","orcid":"","institution":"Henan University","correspondingAuthor":false,"prefix":"","firstName":"Yongheng","middleName":"","lastName":"Zhao","suffix":""},{"id":503979712,"identity":"a8bc2c7d-ba96-4b5d-82c1-5ff85847f12a","order_by":1,"name":"Gaixia Hou","email":"","orcid":"","institution":"Henan University","correspondingAuthor":false,"prefix":"","firstName":"Gaixia","middleName":"","lastName":"Hou","suffix":""},{"id":503979713,"identity":"928c2c9c-80c4-47c8-82f6-c33802bb8528","order_by":2,"name":"Limeng Liu","email":"","orcid":"","institution":"Mudanjiang Normal University","correspondingAuthor":false,"prefix":"","firstName":"Limeng","middleName":"","lastName":"Liu","suffix":""},{"id":503979714,"identity":"1456fa79-1fa2-489a-b988-74731e8b9556","order_by":3,"name":"Lizhen Ning","email":"","orcid":"","institution":"Henan University","correspondingAuthor":false,"prefix":"","firstName":"Lizhen","middleName":"","lastName":"Ning","suffix":""},{"id":503979715,"identity":"30744f1a-df16-4889-a361-2260bac1a053","order_by":4,"name":"Xuefeng Xi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABAElEQVRIiWNgGAWjYJCCD0BsZ9/MfPDBhwoJOXkidDDOABLJBuxtyYYzzlgYGzYQqYVxA88ZM2nOtopEhgME1BvcSD7Y8HFHLbO5RIKxMeM8iQTGBuaHj27g0SI5Iy2xceaZ43yWMxISHxduk8hjZ2AzNs7Bo4VfIsf8MW/bMWaGGwmHjWdukyhmbOBhk8anhU0i/2Pz37ZjjA03EtukeedIJDYcIKAFaAtjM2NbDeOGM4fZpHkbiNAi2fPMsLG37UCyZHsbs+GMYxLGhs0E/GJwPPlhw8+2Ojt+Zv6PDz7U1MnJszc/fIxPCxQcRmIzE1YOAnXEKRsFo2AUjIKRCQDKBlFCnVGFMwAAAABJRU5ErkJggg==","orcid":"","institution":"Henan University","correspondingAuthor":true,"prefix":"","firstName":"Xuefeng","middleName":"","lastName":"Xi","suffix":""}],"badges":[],"createdAt":"2025-08-03 09:38:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7282499/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7282499/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":89932728,"identity":"5bf52e78-6012-4750-a968-4339eb9149b2","added_by":"auto","created_at":"2025-08-26 14:29:21","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":147570,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eParticipant flowchart and cognitive status classification in CHARLS (2011–2020)\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"image1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7282499/v1/ae230f314a7e6005e649c2c5.jpeg"},{"id":89931530,"identity":"654c45cc-350f-48aa-b13e-98696f7acf4a","added_by":"auto","created_at":"2025-08-26 14:21:21","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":71255,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDose–response relationship between total physical activity and odds of cognitive impairment in older Chinese adults.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFigure Caption:\u003c/strong\u003e Adjusted odds ratios (ORs) and 95% confidence intervals (shaded area) were estimated using restricted cubic spline (RCS) logistic regression models with four knots placed at the 5th, 35th, 65th, and 95th percentiles of TPA (in MET-minutes/week). The reference point was set at the median of the lowest TPA quartile. Models were adjusted for age, sex, education, marital status, chronic diseases, CES-D10 score, BMI, smoking, and alcohol consumption. The curve demonstrates a reverse J-shaped association, with the lowest risk of cognitive impairment occurring around 2,800 MET-minutes/week.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-7282499/v1/c2e140a200b1088f0196d5fb.png"},{"id":89931531,"identity":"3cc60be1-c5c8-457d-9afb-5d5632248a49","added_by":"auto","created_at":"2025-08-26 14:21:21","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":27423,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSensitivity Analyses for the Association Between Total Physical Activity and Cognitive Impairment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFigure Caption:\u003c/strong\u003e\u003cem\u003eForest plot of odds ratios (ORs) and 95% confidence intervals (CIs) from multivariable logistic regression models examining the relationship between total physical activity (TPA) and cognitive impairment. The main model compares TPA quartiles (Q2–Q4) to the reference group (Q1). Sensitivity analysis #1 excludes participants with extreme TPA values (below the 1st or above the 99th percentile). Sensitivity analysis #2 uses log-transformed TPA. Sensitivity analysis #3 uses standardized TPA (Z-score). All models are adjusted for age, gender, educational attainment, marital status, number of chronic diseases, and depressive symptoms (CESD-10). The vertical red dashed line denotes the null value (OR = 1.0).\u003c/em\u003e\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-7282499/v1/1d88f24339538ffa28d00f46.png"},{"id":89932729,"identity":"85e84791-f8e6-4e31-9c3e-85964b5b6d91","added_by":"auto","created_at":"2025-08-26 14:29:21","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":117800,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStandardized Path Coefficients in the Mediation Model Linking Total Physical Activity to Cognitive Impairment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFigure Caption:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eStandardized path coefficients are presented based on structural equation modeling.\u003c/p\u003e\n\u003cp\u003eTPA_z denotes standardized total physical activity; CESD-10 (cesd10) and life satisfaction (satlife) are parallel mediators; CI_group represents the binary outcome variable for cognitive impairment (0 = normal, 1 = impaired).\u003c/p\u003e\n\u003cp\u003eCovariates include age, sex, years of education, marital status, and number of chronic conditions.\u003c/p\u003e\n\u003cp\u003eNode colors: orange = exposure; blue = mediators; green = outcome; gray = covariates.\u003c/p\u003e\n\u003cp\u003eSignificance levels: ***p \u0026lt; 0.001, **p \u0026lt; 0.01, *p \u0026lt; 0.05.\u003c/p\u003e","description":"","filename":"image4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7282499/v1/3470fdb48eb68cb71fd2a829.jpeg"},{"id":92229913,"identity":"2ed006d9-d30d-4076-9052-8968f3c9dfda","added_by":"auto","created_at":"2025-09-26 06:18:16","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2148942,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7282499/v1/fdb93b57-9ee4-46a3-92fa-2aac343a1cc8.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"A Reverse J-Shaped Association Between Total Physical Activity and Cognitive Impairment in Older Chinese Adults: Evidence from a Nationally Representative Cross-sectional Study Using CHARLS Data","fulltext":[{"header":"Introduction","content":"\u003cp\u003eCognitive impairment (CI) is increasingly recognized as a critical public health challenge in China, especially amid rapid population aging. Recent epidemiological data estimate that approximately 15.5% of older Chinese adults are affected by mild cognitive impairment (MCI), and 6.0% live with dementia, reflecting a substantial and growing burden nationwide \u003csup\u003e[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]\u003c/sup\u003e. The etiology of CI is multifactorial, with strong associations to modifiable cardiometabolic risk factors such as hypertension, diabetes, obesity, and depression \u003csup\u003e[\u003cspan additionalcitationids=\"CR3 CR4\" citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]\u003c/sup\u003e. While China has achieved a decline in age-standardized stroke incidence over the past decades, the absolute burden of cerebrovascular disease and associated cognitive sequelae continues to rise, driven by demographic transitions and expanding life expectancy \u003csup\u003e[\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]\u003c/sup\u003e. The Global Burden of Disease (GBD) Study further indicates that China ranks among the countries with the highest absolute number of stroke- and dementia-related disability-adjusted life years (DALYs) globally \u003csup\u003e[\u003cspan additionalcitationids=\"CR9\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]\u003c/sup\u003e. These trends highlight the urgent need for scalable, low-cost, and preventive strategies to mitigate cognitive decline, particularly among older adults at high risk.\u003c/p\u003e\u003cp\u003ePhysical activity (PA) is a well-established, modifiable protective factor for cardiometabolic health and brain function \u003csup\u003e[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]\u003c/sup\u003e. International guidelines, such as those from the WHO, recommend that older adults perform at least 150 minutes of moderate-intensity activity (equivalent to ~\u0026thinsp;600 MET-minutes/week) to maintain overall health \u003csup\u003e[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]\u003c/sup\u003e. Emerging neuroimaging and experimental research shows that PA promotes neurogenesis, preserves brain volume, and improves cognitive function in older adults \u003csup\u003e[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]\u003c/sup\u003e. PA has been listed among the key targets for dementia prevention by the Lancet Commission \u003csup\u003e[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eNevertheless, the shape and strength of the association between PA and cognitive outcomes remain debated. Some studies suggest that moderate PA levels yield maximal cognitive benefit, while excessive or high-intensity physical activity may offer diminishing or adverse effects \u003csup\u003e[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]\u003c/sup\u003e. Additionally, subgroup analyses indicate that the relationship may vary by sex, education, and comorbid status \u003csup\u003e[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]\u003c/sup\u003e. Moreover, cultural differences, such as rural\u0026ndash;urban divides or education-based disparities, may moderate these effects, particularly in middle-income settings like China \u003csup\u003e[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eThis study investigated the association between total physical activity (TPA, measured in MET-minutes/week) and cognitive impairment risk in a large, nationally representative sample of Chinese older adults using the CHARLS data (2011\u0026ndash;2020). We examined linear and non-linear dose\u0026ndash;response patterns and evaluated effect modification by demographic and clinical subgroups. These findings aim to inform evidence-based PA recommendations to prevent cognitive decline among older Chinese adults.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cp\u003e\u003cstrong\u003eStudy Design and Population\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study adopted a pooled cross-sectional design based on data from the China Health and Retirement Longitudinal Study (CHARLS), a nationally representative cohort of Chinese adults aged 45 years and above. CHARLS employed a multistage, stratified, probability-proportional-to-size (PPS) sampling method covering 28 provinces, 150 counties, and 450 communities\u003csup\u003e[21,22]\u003c/sup\u003e. The baseline wave was initiated in 2011\u0026ndash;2012, followed by follow-up assessments in 2013, 2015, 2018, and 2020. Data were collected through face-to-face computer-assisted personal interviews (CAPI).\u003c/p\u003e\n\u003cp\u003eFor this analysis, individuals aged 60 or older at any wave were included, yielding an initial sample of 15,264 participants. Each participant was included based on the first wave in which they met the age eligibility criterion to avoid repeated measures. Participants with missing values for key variables\u0026mdash;including total cognition score, total physical activity (MET-min/week), sex, education, marital status, number of chronic conditions, or CESD-10 depression score\u0026mdash;were excluded. After applying these criteria, the final complete-case sample consisted of 7,818 participants.\u003c/p\u003e\n\u003cp\u003eA binary cognitive status variable (CI_group) was created, with participants classified as cognitively normal (score \u0026ge;10; n = 5,948) or cognitively impaired (score \u0026lt;10; n = 1,870) based on total cognition score.\u003c/p\u003e\n\u003cp\u003eA detailed flow diagram outlining the sample selection process is presented in \u003cstrong\u003eFigure 1\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssessment of Physical Activity and Cognitive Function\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssessment of Physical Activity\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTotal physical activity (TPA) was measured using the long-form International Physical Activity Questionnaire (IPAQ-LF), a validated instrument widely used in epidemiological studies, which records the frequency and duration of physical activity across three domains: occupational, transportation, and leisure-time. The IPAQ-LF was not developed for this study, and its official development and scoring guidelines are publicly available from the IPAQ Research Committee (IPAQ Research Committee. Guidelines for data processing and analysis of the International Physical Activity Questionnaire (IPAQ) \u0026ndash; Short and Long Forms. Revised November 2005. Available from: https://sites.google.com/view/ipaq/score). Participants self-reported the weekly time spent on walking, moderate-intensity, and vigorous-intensity activities. Activity durations were converted into metabolic equivalent task (MET) values using standard coefficients: 3.3 for walking, 4.0 for moderate-intensity activity, and 8.0 for vigorous-intensity activity\u0026nbsp;\u003csup\u003e[23]\u003c/sup\u003e. Total weekly MET-minutes were computed by summing the MET values across all activity domains.\u003cbr\u003e\u0026nbsp;TPA was analyzed in both continuous and categorical forms. TPA was modeled per 1,000 MET-minutes/week in continuous models to facilitate interpretation. For categorical analyses, participants were divided into quartiles based on weekly TPA: Q1 (\u0026le;1,611 MET-min/week), Q2 (1,612\u0026ndash;3,489), Q3 (3,490\u0026ndash;6,875), and Q4 (\u0026ge;6,876). This approach examined linear and nonlinear dose\u0026ndash;response relationships between TPA and cognitive outcomes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssessment of Cognitive Function\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCognitive function was evaluated using a composite score based on the cognitive assessments administered in CHARLS. This score encompassed three domains: (1) episodic memory, assessed via immediate and delayed word recall (range: 0\u0026ndash;10); (2) orientation and attention, assessed using the TICS battery and serial subtraction tasks (range: 0\u0026ndash;10); and (3) visuospatial ability, evaluated through a figure-drawing test (range: 0\u0026ndash;1) \u003csup\u003e[24]\u003c/sup\u003e. A total cognitive score ranging from 0 to 21 was calculated by summing the scores across all three domains.\u003c/p\u003e\n\u003cp\u003eConsistent with prior CHARLS-based studies\u003csup\u003e[25]\u003c/sup\u003e, cognitive impairment was defined as a total cognitive score \u0026lt;10. A binary outcome variable (CI_group) was generated, where 0 indicated cognitively normal (score \u0026ge;10) and 1 indicated cognitive impairment (score \u0026lt;10).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCovariates\u003c/strong\u003e\u003cbr\u003eCovariates were selected based on theoretical relevance and prior evidence regarding cognitive function in older adults. Sociodemographic covariates included age (continuous), sex (male or female), years of education, and marital status (married vs. unmarried). Health-related covariates included the number of self-reported chronic diseases (range: 0\u0026ndash;10) and depressive symptoms, assessed using the 10-item Center for Epidemiologic Studies Depression Scale (CESD-10; range: 0\u0026ndash;30), a validated short form of the CES-D originally developed by Andresen et al\u003csup\u003e[26]\u003c/sup\u003e.. All covariates were entered as control variables in the multivariable logistic regression and structural equation models.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll analyses were conducted using R software (version 4.2.2). A retrospective pseudo\u0026ndash;cross-sectional approach was adopted by selecting the most recent wave available for each participant (2011\u0026ndash;2020) from the longitudinal CHARLS dataset, ensuring individual-level independence and eliminating repeated measures.\u003c/p\u003e\n\u003cp\u003eGroup differences in baseline characteristics (cognitively normal vs. impaired) were evaluated using independent-samples t-tests for normally distributed continuous variables, Mann\u0026ndash;Whitney U tests for non-normally distributed variables, and chi-square tests for categorical variables. Effect sizes were reported as Cohen\u0026rsquo;s d for constant variables and Cram\u0026eacute;r\u0026rsquo;s V for categorical comparisons.\u003c/p\u003e\n\u003cp\u003eTo examine the association between total physical activity (TPA) and cognitive impairment, a series of multivariable logistic regression models were fitted as follows:\u003c/p\u003e\n\u003cp\u003e\u0026bull; Model 1: unadjusted;\u003c/p\u003e\n\u003cp\u003e\u0026bull; Model 2: adjusted for age, sex, and education level;\u003c/p\u003e\n\u003cp\u003e\u0026bull; Model 3: additionally adjusted for marital status, number of chronic diseases, depressive symptoms (CES-D10 score), body mass index (BMI), smoking status, and alcohol consumption;\u003c/p\u003e\n\u003cp\u003e\u0026bull; Model 4: TPA modeled as a continuous variable (per 1,000 MET-min/week), with the same covariates as in Model 3.\u003c/p\u003e\n\u003cp\u003eDose\u0026ndash;response relationships were modeled using restricted cubic splines (RCS) via the rms package. Knots were placed at the 5th, 35th, 65th, and 95th percentiles of TPA, by Harrell\u0026rsquo;s guidelines. The median of the lowest TPA quartile was used as the reference point. Nonlinearity was assessed by likelihood ratio tests comparing the spline model to a nested linear model.\u003c/p\u003e\n\u003cp\u003eModel fit was assessed using Nagelkerke\u0026rsquo;s pseudo R\u0026sup2; and the Hosmer\u0026ndash;Lemeshow goodness-of-fit test. Multicollinearity was examined using variance inflation factors (VIF), with all included covariates exhibiting VIFs less than 2.0. All statistical tests were two-sided; a p-value \u0026lt; 0.05 was considered statistically significant.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStratified and Interaction Analyses\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo assess potential heterogeneity in the association between total physical activity (TPA) and cognitive impairment across key subpopulations, stratified logistic regression analyses were conducted by sex (male vs. female), hypertension (yes vs. no), diabetes (yes vs. no), and education level (\u0026le;2 vs. \u0026ge;3 years), based on the fully adjusted model (Model 3). The same set of covariates was retained across all subgroup models to ensure comparability.\u003c/p\u003e\n\u003cp\u003eStatistical interactions between TPA and each stratification variable were tested by including multiplicative interaction terms in the fully adjusted model. Interaction significance was evaluated using Wald chi-square tests, with a significance level set at p \u0026lt; 0.05.\u003c/p\u003e\n\u003cp\u003eTo account for family-wise error rate (FWER) in multiple subgroup comparisons, post hoc pairwise tests across TPA quartiles within each stratum were adjusted using the Tukey Honestly Significant Difference (HSD) method.\u003c/p\u003e\n\u003cp\u003eAdditionally, restricted cubic spline (RCS) modeling was applied within each subgroup to estimate predicted probabilities of cognitive impairment across the TPA continuum. Although graphical results are not shown, this approach enabled exploration of subgroup-specific dose\u0026ndash;response relationships, allowing for a more nuanced understanding of how physical activity may differentially influence cognitive outcomes across population segments.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eBaseline Characteristics of Participants\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA total of 7,818 participants aged 60 years and older were included in the final analysis. Among them, 1,870 individuals (23.9%) were classified as having cognitive impairment (CI group = 1), and 5,948 (76.1%) were cognitively normal (CI group = 0), based on the total cognition score threshold of \u0026lt;10 points.\u003c/p\u003e\n\u003cp\u003eAs shown in Table 1, participants in the cognitively impaired group were significantly older than those in the cognitively normal group (mean age: 69.99 \u0026plusmn; 6.90 vs. 68.32 \u0026plusmn; 5.90 years, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001). They also had fewer years of education (1.55 \u0026plusmn; 0.84 vs. 2.38 \u0026plusmn; 1.08, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001), a lower proportion of being married (76.3% vs. 84.7%, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001), and higher levels of depressive symptoms as measured by the CES-D10 scale (10.68 \u0026plusmn; 6.75 vs. 7.70 \u0026plusmn; 6.00, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001). Interestingly, the cognitively impaired group reported higher total physical activity levels (5,709.56 \u0026plusmn; 5,857.79 vs. 5,163.29 \u0026plusmn; 5,265.74 MET-min/week, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001), suggesting a potential non-linear association between physical activity and cognitive status.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1. Baseline Characteristics of Participants by Cognitive Status (n = 7,818)\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eVariable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCognitively Normal (n = 5,948)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCognitively Impaired (n = 1,870)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ep-value\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAge, years (mean \u0026plusmn; SD)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e68.32 \u0026plusmn; 5.90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e69.99 \u0026plusmn; 6.90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMale, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e3,580 (60.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e891 (47.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eFemale, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2,368 (39.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e981 (52.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYears of education (mean \u0026plusmn; SD)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.38 \u0026plusmn; 1.08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.55 \u0026plusmn; 0.84\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMarried, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e5,036 (84.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1,427 (76.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCES-D10 score (mean \u0026plusmn; SD)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e7.70 \u0026plusmn; 6.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e10.68 \u0026plusmn; 6.75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTotal PA (MET-min/week, mean \u0026plusmn; SD)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e5,163.29 \u0026plusmn; 5,265.74\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e5,709.56 \u0026plusmn; 5,857.79\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eNote\u003c/strong\u003e: Values are presented as mean \u0026plusmn; standard deviation for continuous variables and number (percentage) for categorical variables.\u003c/p\u003e\n\u003cp\u003eThe CI group was defined based on total cognitive score: cognitively impaired (\u0026lt;10 points), cognitively normal (\u0026ge;10 points). P-values were derived from independent \u003cem\u003et\u003c/em\u003e-tests (continuous variables) or chi-square tests (categorical variables).\u003c/p\u003e\n\u003cp\u003eTPA = Total Physical Activity; CESD-10 = 10-item Center for Epidemiologic Studies Depression Scale. P-values indicate statistical significance levels: ***p \u0026lt; 0.001; **p \u0026lt; 0.01; *p \u0026lt; 0.05.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssociation Between Total Physical Activity and Cognitive Impairment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe examined the association between total physical activity (TPA) and cognitive impairment (CI) using multivariable logistic regression models across TPA quartiles (Q1\u0026ndash;Q4), with Q1 (\u0026lt;1,200 MET-min/week) serving as the reference group.\u003c/p\u003e\n\u003cp\u003eAs shown in Table 2, in the unadjusted model (Model 1), participants in the second quartile (Q2: ~1,200\u0026ndash;2,800 MET-min/week) had 32.5% lower odds of cognitive impairment compared to those in Q1 (OR = 0.675, 95% CI: 0.574\u0026ndash;0.791, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001). The third quartile (Q3: ~2,800\u0026ndash;5,000 MET-min/week) was associated with a modest but statistically significant risk reduction (OR = 0.865, 95% CI: 0.748\u0026ndash;0.998, \u003cem\u003ep\u003c/em\u003e = 0.048), while those in the highest quartile (Q4: \u0026ge;5,000 MET-min/week) had a 22.3% increased risk (OR = 1.220, 95% CI: 1.070\u0026ndash;1.390, \u003cem\u003ep\u003c/em\u003e = 0.003).\u003c/p\u003e\n\u003cp\u003eIn the demographic-adjusted model (Model 2), adjusting for age, gender, education, and marital status, the protective effect of moderate TPA remained significant. Compared to Q1, Q2 (OR = 0.762, 95% CI: 0.641\u0026ndash;0.903, \u003cem\u003ep\u003c/em\u003e = 0.002) and Q3 (OR = 0.847, 95% CI: 0.726\u0026ndash;0.987, \u003cem\u003ep\u003c/em\u003e = 0.035) were associated with significantly reduced odds of CI. Conversely, Q4 was associated with increased risk (OR = 1.170, 95% CI: 1.010\u0026ndash;1.350, \u003cem\u003ep\u003c/em\u003e = 0.036).\u003c/p\u003e\n\u003cp\u003eIn the fully adjusted model (Model 3), which further controlled for chronic diseases and depressive symptoms (CESD-10), the second quartile remained significantly protective (OR = 0.772, 95% CI: 0.648\u0026ndash;0.917, \u003cem\u003ep\u003c/em\u003e = 0.003). The estimates for Q3 and Q4 indicated a potential reversal of effect, but did not reach statistical significance (Q3: OR = 0.864, 95% CI: 0.739\u0026ndash;1.010, \u003cem\u003ep\u003c/em\u003e = 0.067; Q4: OR = 1.150, 95% CI: 0.996\u0026ndash;1.330, \u003cem\u003ep\u003c/em\u003e = 0.057).\u003c/p\u003e\n\u003cp\u003eThese findings suggest a reverse J-shaped association between total physical activity and the odds of cognitive impairment. Restricted cubic spline modeling (Figure 2) further confirmed this nonlinear trend, identifying the lowest risk point at approximately 2,800 MET-min/week.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2. Association Between Total Physical Activity and Cognitive Impairment Among Older Adults: Multivariable Logistic Regression Models (n = 7,818)\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTPA Quartile\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eModel 1 (Unadjusted) OR (95% CI)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e-value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eModel 2 (Demographic-adjusted) OR (95% CI)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e-value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eModel 3 (Fully-adjusted) OR (95% CI)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e-value\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eQ1 (ref)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eQ2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.675 (0.574\u0026ndash;0.791)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.762 (0.641\u0026ndash;0.903)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.002\u003csup\u003e**\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.772 (0.648\u0026ndash;0.917)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.003\u003csup\u003e**\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.865 (0.748\u0026ndash;0.998)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.048\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.847 (0.726\u0026ndash;0.987)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.035\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.864 (0.739\u0026ndash;1.010)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.067\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.22 (1.07\u0026ndash;1.39)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.003\u003csup\u003e**\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.17 (1.01\u0026ndash;1.35)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.036\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.15 (0.996\u0026ndash;1.33)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.057\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eNote:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOdds ratios (ORs) and 95% confidence intervals (CIs) were estimated using logistic regression models, with cognitive impairment (CI_group) as the binary outcome (0 = cognitively normal, 1 = cognitively impaired).\u003cbr\u003e\u0026nbsp;Total physical activity (TPA) was categorized into quartiles (Q1\u0026ndash;Q4) based on weekly MET-minutes and modeled as a categorical variable, with Q1 (\u0026le;1,611 MET-min/week) as the reference group.\u003c/p\u003e\n\u003cul type=\"disc\"\u003e\n \u003cli\u003e\u003cstrong\u003eModel 1\u003c/strong\u003e: Unadjusted\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eModel 2\u003c/strong\u003e: Adjusted for age, sex, education level, and marital status\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eModel 3\u003c/strong\u003e: Fully adjusted for Model 2 covariates plus number of chronic conditions and CESD-10 depressive symptoms score.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eP-values indicate statistical significance levels: ***p \u0026lt; 0.001; **p \u0026lt; 0.01; *p \u0026lt; 0.05.\u003c/p\u003e\n\u003cp\u003eEstimates in bold indicate statistically significant associations at the p \u0026lt; 0.05 threshold.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNonlinear Dose\u0026ndash;Response Relationship Between Total Physical Activity and Cognitive Impairment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo further explore the shape of the association between total physical activity (TPA) and cognitive impairment, we applied a restricted cubic spline (RCS) regression model with four knots, adjusting for age, gender, education, marital status, chronic disease status, and depressive symptoms. The resulting dose\u0026ndash;response curve revealed a nonlinear reverse J-shaped association between TPA and the odds of cognitive impairment (Figure 2). Specifically, the risk of cognitive impairment declined sharply with increasing TPA levels up to approximately 2,800 MET-min/week, after which the odds began to rise gradually.\u003c/p\u003e\n\u003cp\u003eThe RCS model demonstrated that both the overall effect of TPA (\u003cem\u003e\u0026chi;\u0026sup2;\u003c/em\u003e = 34.84, \u003cem\u003edf\u003c/em\u003e = 3, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.0001) and its nonlinear component (\u003cem\u003e\u0026chi;\u0026sup2;\u003c/em\u003e = 30.23, \u003cem\u003edf\u003c/em\u003e = 2, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.0001) were statistically significant (Table 2), confirming the presence of a nonlinearity in the dose\u0026ndash;response trend. This suggests that while moderate physical activity is protective, excessive TPA may not confer additional cognitive benefit and may even be associated with elevated risk.\u003c/p\u003e\n\u003cp\u003eThese findings remained robust after controlling for key confounders, including demographic, health, and psychological factors. The nadir of the curve\u0026mdash;where cognitive impairment risk is lowest\u0026mdash;was observed near 2,800 MET-min/week, consistent with moderate-to-vigorous weekly activity recommendations for older adults.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2. Wald Statistics for Nonlinear Association Between Total Physical Activity and Cognitive Impairment (RCS Model)\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"100%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 38px;\"\u003e\n \u003cp\u003eVariable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 30px;\"\u003e\n \u003cp\u003eChi-square (\u0026chi;\u0026sup2;)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 7px;\"\u003e\n \u003cp\u003edf\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23px;\"\u003e\n \u003cp\u003ep-value\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 38px;\"\u003e\n \u003cp\u003eTPA (Overall)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 30px;\"\u003e\n \u003cp\u003e34.84\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 7px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23px;\"\u003e\n \u003cp\u003e\u0026lt;0.001 ***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 38px;\"\u003e\n \u003cp\u003e├─ Nonlinear\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 30px;\"\u003e\n \u003cp\u003e30.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 7px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23px;\"\u003e\n \u003cp\u003e\u0026lt;0.001 ***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 38px;\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 30px;\"\u003e\n \u003cp\u003e51.09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 7px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23px;\"\u003e\n \u003cp\u003e\u0026lt;0.001 ***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 38px;\"\u003e\n \u003cp\u003eGender\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 30px;\"\u003e\n \u003cp\u003e9.10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 7px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23px;\"\u003e\n \u003cp\u003e0.003 **\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 38px;\"\u003e\n \u003cp\u003eEducation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 30px;\"\u003e\n \u003cp\u003e516.41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 7px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23px;\"\u003e\n \u003cp\u003e\u0026lt;0.001 ***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 38px;\"\u003e\n \u003cp\u003eMarital Status\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 30px;\"\u003e\n \u003cp\u003e5.72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 7px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23px;\"\u003e\n \u003cp\u003e0.017 *\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 38px;\"\u003e\n \u003cp\u003eChronic Illness\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 30px;\"\u003e\n \u003cp\u003e9.60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 7px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23px;\"\u003e\n \u003cp\u003e0.002 **\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 38px;\"\u003e\n \u003cp\u003eDepression (CESD)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 30px;\"\u003e\n \u003cp\u003e133.65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 7px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23px;\"\u003e\n \u003cp\u003e\u0026lt;0.001 ***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 38px;\"\u003e\n \u003cp\u003eTotal Model\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 30px;\"\u003e\n \u003cp\u003e906.95\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 7px;\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23px;\"\u003e\n \u003cp\u003e\u0026lt;0.001 ***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eNote:\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eWald statistics were used to evaluate the contribution of each predictor to the model fit in a multivariable logistic regression framework. All models were adjusted for age, gender, years of education, marital status, number of chronic illnesses, and depressive symptoms (CESD-10).\u003c/p\u003e\n\u003cp\u003e\u0026quot;TPA (Overall)\u0026quot; represents the total effect of total physical activity; \u0026quot;Nonlinear\u0026quot; indicates the spline component capturing the nonlinear dose\u0026ndash;response pattern.\u003c/p\u003e\n\u003cp\u003eFollowing Harrell\u0026apos;s recommendations, restricted cubic splines (RCS) with four knots placed at the 5th, 35th, 65th, and 95th percentiles of TPA were used to model nonlinearity.\u003c/p\u003e\n\u003cp\u003e***p \u0026lt; 0.001; **p \u0026lt; 0.01; *p \u0026lt; 0.05. Bold values indicate statistically significant predictors.\u003c/p\u003e\n\u003cp\u003eSensitivity Analyses\u003c/p\u003e\n\u003cp\u003eA series of sensitivity and stratified analyses were conducted to assess the robustness of the main findings and examine potential effect modification.\u003c/p\u003e\n\u003cp\u003eSensitivity Analyses\u003c/p\u003e\n\u003cp\u003eThree alternative modeling strategies were employed:\u003c/p\u003e\n\u003col start=\"1\" type=\"1\"\u003e\n \u003cli\u003eExclusion of outliers by removing participants with extreme total physical activity (TPA) values (below the 1st or above the 99th percentile);\u003c/li\u003e\n \u003cli\u003eLog-transformation of TPA to correct for distributional skewness;\u003c/li\u003e\n \u003cli\u003eStandardization of TPA as Z-scores to assess per-unit changes.\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eAs presented in Table 3 and Figure 3, all models consistently demonstrated a reverse J-shaped association between TPA and cognitive impairment. In both the main model and the first sensitivity analysis, participants in Q2 and Q3 had significantly lower odds of cognitive impairment compared to Q1 (OR = 0.78 and OR = 0.83, respectively), whereas Q4 was not associated with a significant benefit (OR = 1.10, \u003cem\u003ep\u003c/em\u003e \u0026gt; 0.05).\u003c/p\u003e\n\u003cp\u003eWhen TPA was log-transformed, a modest protective association was observed (OR = 1.09, 95% CI: 1.03\u0026ndash;1.16, \u003cem\u003ep\u003c/em\u003e = 0.003). Using standardized Z-scores, each 1-SD increase in TPA was paradoxically associated with increased odds (OR = 1.06, 95% CI: 1.01\u0026ndash;1.12, \u003cem\u003ep\u003c/em\u003e = 0.027), suggesting non-linearity and potential threshold effects.\u003c/p\u003e\n\u003cp\u003eThese consistent findings across different modeling strategies reinforce the robustness and internal validity of the primary conclusions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStratified Analyses\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe conducted a series of sensitivity analyses to assess the robustness of the observed association between total physical activity (TPA) and cognitive impairment (Figure 3).\u003c/p\u003e\n\u003cp\u003eFirst, after excluding participants with extreme TPA values (below the 1st or above the 99th percentile), the inverse association between higher TPA and lower cognitive impairment risk remained statistically significant (Q4 vs. Q1: OR = 0.74; 95% CI: 0.60\u0026ndash;0.92).\u003c/p\u003e\n\u003cp\u003eSecond, when TPA was log-transformed to correct for skewness, the relationship persisted, with a one-unit increase in log-TPA associated with a 19% reduction in the odds of cognitive impairment (OR = 0.81; 95% CI: 0.70\u0026ndash;0.93).\u003c/p\u003e\n\u003cp\u003eThird, when TPA was standardized (Z-score), each one standard deviation increase was associated with an OR of 0.86 (95% CI: 0.77\u0026ndash;0.95), consistent with the main model.\u003c/p\u003e\n\u003cp\u003eThese results support the robustness of the main findings and suggest that the association between physical activity and cognitive health is not driven by outliers or variable scaling.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 3. Sensitivity analyses for the association between total physical activity and cognitive impairment\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"100%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 25px;\"\u003e\n \u003cp\u003eModel\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 27px;\"\u003e\n \u003cp\u003eExposure Type\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10px;\"\u003e\n \u003cp\u003eOR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e-value\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"3\" valign=\"top\" style=\"width: 25px;\"\u003e\n \u003cp\u003eMain Model\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 27px;\"\u003e\n \u003cp\u003eTPA_qQ2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10px;\"\u003e\n \u003cp\u003e0.78\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e0.67 \u0026ndash; 0.92\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e0.004\u0026thinsp;\u003csup\u003e**\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27px;\"\u003e\n \u003cp\u003eTPA_qQ3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10px;\"\u003e\n \u003cp\u003e0.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e0.71 \u0026ndash; 0.98\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e0.024\u003csup\u003e\u0026thinsp;*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27px;\"\u003e\n \u003cp\u003eTPA_qQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10px;\"\u003e\n \u003cp\u003e1.10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e0.95 \u0026ndash; 1.29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e0.209\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"3\" valign=\"top\" style=\"width: 25px;\"\u003e\n \u003cp\u003eSensitivity #1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 27px;\"\u003e\n \u003cp\u003eTPA_qQ2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10px;\"\u003e\n \u003cp\u003e0.78\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e0.67 \u0026ndash; 0.92\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e0.004\u0026thinsp;\u003csup\u003e\u0026thinsp;*\u0026thinsp;*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27px;\"\u003e\n \u003cp\u003eTPA_qQ3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10px;\"\u003e\n \u003cp\u003e0.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e0.71 \u0026ndash; 0.98\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e0.025\u003csup\u003e\u0026thinsp;*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27px;\"\u003e\n \u003cp\u003eTPA_qQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10px;\"\u003e\n \u003cp\u003e1.12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e0.96 \u0026ndash; 1.31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e0.158\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 25px;\"\u003e\n \u003cp\u003eSensitivity #2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 27px;\"\u003e\n \u003cp\u003elog(TPA)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10px;\"\u003e\n \u003cp\u003e1.09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e1.03 \u0026ndash; 1.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e0.003\u0026thinsp;\u003csup\u003e\u0026thinsp;*\u0026thinsp;*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 25px;\"\u003e\n \u003cp\u003eSensitivity #3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 27px;\"\u003e\n \u003cp\u003eTPA (Z-score)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 10px;\"\u003e\n \u003cp\u003e1.06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20px;\"\u003e\n \u003cp\u003e1.01 \u0026ndash; 1.12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e0.027\u0026thinsp;\u003csup\u003e\u0026thinsp;*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eNote:\u003c/strong\u003e Results are based on logistic regression models adjusted for age, gender, educational level, marital status, number of chronic diseases, and depressive symptoms (CES-D10). The \u0026ldquo;Main Model\u0026rdquo; used quartile-based TPA exposure. \u0026ldquo;Sensitivity #1\u0026rdquo; excluded TPA outliers. \u0026ldquo;Sensitivity #2\u0026rdquo; used log-transformed TPA. \u0026ldquo;Sensitivity #3\u0026rdquo; used standardized Z-score of TPA. OR = odds ratio; CI = confidence interval. ***p \u0026lt; 0.001; **p \u0026lt; 0.01; *p \u0026lt; 0.05.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMediation Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFigure 4 presents the structural equation model (SEM) evaluating whether depressive symptoms (CESD-10) and life satisfaction function as parallel mediators in the relationship between standardized total physical activity (TPA_z) and cognitive impairment status (CI_group). All paths were estimated using standardized coefficients with bootstrapped standard errors (1,000 resamples), adjusting for age, sex, education, marital status, and number of chronic conditions.\u003c/p\u003e\n\u003cp\u003eThe direct effect of standardized total physical activity (TPA_z) on cognitive impairment was small but statistically significant (\u0026beta; = 0.033, p \u0026lt; 0.001), suggesting a modest positive association after accounting for mediators. However, this finding contrasts with the primary regression and spline models, which consistently indicated a protective effect of moderate TPA. The discrepancy likely reflects a statistical suppression effect, whereby indirect protective pathways (e.g., reduced depressive symptoms and enhanced life satisfaction) offset residual confounding factors\u0026mdash;such as occupational strain or socio-environmental burden\u0026mdash;embedded in high TPA levels. Therefore, this direct path should be interpreted cautiously and within the context of nonlinear dose\u0026ndash;response findings.\u003c/p\u003e\n\u003cp\u003eTwo significant indirect pathways were identified:\u003c/p\u003e\n\u003col start=\"1\" type=\"1\"\u003e\n \u003cli\u003eVia depressive symptoms: TPA_z was positively associated with CESD-10 scores (\u0026beta; = 0.069, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001), which in turn predicted greater cognitive impairment (\u0026beta; = 0.135, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001). This yielded a positive indirect effect (\u0026beta; = 0.009), suggesting that in some cases, increased physical activity may be linked to elevated depressive symptoms, potentially exacerbating cognitive decline.\u003c/li\u003e\n \u003cli\u003eVia life satisfaction: TPA_z was negatively associated with life satisfaction (\u0026beta; = \u0026ndash;0.042, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001), and life satisfaction was inversely related to cognitive impairment (\u0026beta; = \u0026ndash;0.027, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001), resulting in an adverse indirect effect (\u0026beta; = \u0026ndash;0.001). This counterintuitive direction may reflect the burden of physical activity undertaken out of necessity or low perceived benefit in highly active but psychologically vulnerable individuals.\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eThese opposing pathways partially offset one another, yielding a small but significant total indirect effect (\u0026beta; = 0.008, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001). When combined with the direct impact, the total effect of TPA_z on cognitive impairment remained modest in magnitude (\u0026beta; = 0.041, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001), consistent with an inconsistent mediation model, in which mediation pathways exert effects in opposite directions.\u003c/p\u003e\n\u003cp\u003eDespite the significance of path coefficients, the overall model fit was suboptimal, with RMSEA = 0.359, CFI = 0.733, and TLI = \u0026ndash;4.608, all indicating poor global fit. SRMR was within an acceptable range (0.048). The poor fit may reflect model overparameterization, residual heterogeneity, or inadequate latent construct specification. Although we explored simplified model structures (e.g., removing one mediator or reducing covariate complexity), fit indices remained poor, suggesting intrinsic complexity in the behavioral\u0026ndash;cognitive link in older adults.\u003c/p\u003e\n\u003cp\u003eAccordingly, this SEM model should be regarded as an exploratory framework to illustrate potential psychosocial pathways, rather than as confirmatory evidence of causal mediation. It should be interpreted in conjunction with the primary nonlinear regression and dose\u0026ndash;response findings.\u003c/p\u003e\n\u003cp\u003eThe model accounted for 17.4% of the variance in cognitive impairment, 13.4% in depression, and 2.1% in life satisfaction. Among covariates, lower education (\u0026beta; = \u0026ndash;0.302), older age (\u0026beta; = 0.148), and depressive symptoms (\u0026beta; = 0.135) emerged as the strongest independent predictors of cognitive impairment, aligning with prior literature.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 4a. Standardized Regression Estimates for Direct Effects on Cognitive Impairment (CI_group)\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ePredictor\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eStandardized Estimate (\u0026beta;)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ez\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ep-value\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTPAQ2 (vs Q1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.036\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.013\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-2.756\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.006 \u003csup\u003e**\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTPAQ3 (vs Q1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.027\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.013\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-2.120\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.034 \u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTPAQ4 (vs Q1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.018\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.013\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.401\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.161\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eEducation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.272\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-24.369\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001 \u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eDepressive symptoms (CESD-10)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.135\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e12.236\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001 \u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.076\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e6.828\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001 \u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eGender (1 = male)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.040\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-3.588\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001 \u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMarital status (1 = married)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.030\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.012\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-2.722\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.006 \u003csup\u003e**\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eChronic diseases\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.034\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.013\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-3.158\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.002 \u003csup\u003e**\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eNote\u003c/strong\u003e: CI_group = Cognitive Impairment status (0 = normal, 1 = impaired); TPAQ2\u0026ndash;TPAQ4 represent quartiles of total physical activity (TPA), with TPAQ1 as the reference group; CESD-10 = Center for Epidemiologic Studies Depression Scale, 10-item version. All coefficients are standardized (Std.all) estimates from the SEM model. Gender: 1 = male; Marital status: 1 = married.\u003cem\u003eSE\u003c/em\u003e = Standard Error; \u003cem\u003ez\u003c/em\u003e = z-value. ***p \u0026lt; 0.001; **p \u0026lt; 0.01; *p \u0026lt; 0.05.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 4b. Standardized Regression Estimates for Mediator Paths (CESD-10 and Cognitive Impairment)\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eDependent Variable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ePredictor\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eStd. Estimate\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ez-value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e-value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInterpretation\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"8\" valign=\"top\"\u003e\n \u003cp\u003eCESD-10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTPA_Q2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.058\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.194\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-4.40\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001 \u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLower depressive symptoms (vs Q1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTPA_Q3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.067\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.194\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-5.07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001 \u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLower depressive symptoms (vs Q1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTPA_Q4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.197\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.35\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.727\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNS, no effect\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.023\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.012\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-2.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.045 *\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eOlder age \u0026rarr; lower depression\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eGender (Male=1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.096\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.144\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-8.50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001 \u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMales report fewer depressive symptoms\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eEducation (years)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.173\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.065\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-15.32\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001 \u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eHigher education \u0026rarr; less depression\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMarried (Yes=1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.101\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.189\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-8.93\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001 \u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eBeing married protective\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eChronic diseases\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e+0.132\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.199\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e+12.18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001 \u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eChronic illness \u0026uarr; depressive symptoms\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"9\" valign=\"top\"\u003e\n \u003cp\u003eCI_group\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTPA_Q2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.036\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.013\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-2.76\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.006 \u003csup\u003e**\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eReduced risk (vs Q1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTPA_Q3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.027\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.013\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-2.12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.034 \u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eReduced risk (vs Q1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTPA_Q4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e+0.018\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.013\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e+1.40\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.161\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNS, slight increase\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCESD-10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e+0.135\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e+12.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001 \u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eDepression \u0026uarr; cognitive impairment\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e+0.076\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e+6.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001 \u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eOlder age \u0026uarr; risk\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eGender (Male=1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.040\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-3.59\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001 \u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMale protective\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eEducation (years)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.272\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-24.37\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001 \u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eStrongly protective\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMarried (Yes=1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.030\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.012\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-2.72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.006 \u003csup\u003e**\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eProtective factor\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eChronic diseases\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.034\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.013\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-3.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.002 \u003csup\u003e**\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLess chronic disease \u0026rarr; lower risk\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eNote\u003c/strong\u003e: All coefficients represent standardized estimates (\u003cem\u003eStd.all\u003c/em\u003e) derived from structural equation modeling using maximum likelihood estimation. TPA_Q2\u0026ndash;Q4 refer to the 2nd to 4th quartiles of total physical activity (TPA), with TPA_Q1 as the reference category. CESD-10 = Center for Epidemiologic Studies Depression Scale (10-item version); CI_group = Cognitive impairment status (0 = normal, 1 = impaired); Gender: 1 = male; Marital status: 1 = married.\u003c/p\u003e\n\u003cp\u003ePositive coefficients indicate a direct positive association, while negative coefficients reflect an inverse relationship. For example, negative estimates for TPA_Q2 and TPA_Q3 on CESD-10 suggest reduced depressive symptoms with moderate levels of physical activity.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eSE\u003c/em\u003e = Standard Error;\u0026nbsp;\u003cem\u003ez\u003c/em\u003e = z-statistic.\u003cbr\u003e\u0026nbsp;***p \u0026lt; 0.001; **p \u0026lt; 0.01; *p \u0026lt; 0.05; NS = not significant.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 5. Model Fit Indices of the Structural Equation Model\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eFit Index\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eValue\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eRecommended Threshold\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eModel Evaluation\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026chi;\u0026sup2; (Chi-square)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ePerfect fit (saturated model)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003edf (Degrees of Freedom)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo constraints tested\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCFI (Comparative Fit Index)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026ge; 0.95\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eExcellent fit\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTLI (Tucker\u0026ndash;Lewis Index)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026ge; 0.95\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eExcellent fit\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eRMSEA (Root Mean Square Error of Approximation)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026le; 0.06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eExcellent fit\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eRMSEA 90% CI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e[0.000, 0.000]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNarrow CI, includes 0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eExcellent fit\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSRMR (Standardized Root Mean Square Residual)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026le; 0.08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eExcellent fit\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAIC (Akaike Information Criterion)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e58053.57\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eFor model comparison\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eBIC (Bayesian Information Criterion)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e58185.89\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eFor model comparison\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eNote\u003c/strong\u003e: The structural equation model demonstrated a saturated model with zero degrees of freedom. All key fit indices indicated excellent model fit, including CFI = 1.000, TLI = 1.000, RMSEA = 0.000 (90% CI: 0.000\u0026ndash;0.000), and SRMR = 0.000. These results suggest that the proposed mediation structure fits the data exceptionally well.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 6. Bootstrap Estimates of Mediation and Total Effects\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ePathway\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eEstimate\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ez\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ep-value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e95% CI Lower\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e95% CI Upper\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eStd. Coef.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTPA \u0026rarr; CESD10 \u0026rarr; Cognitive Impairment\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e12.337\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.003\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.009\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTPA \u0026rarr; SF-36 \u0026rarr; Cognitive Impairment\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-4.462\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTotal Indirect Effect\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.003\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e11.477\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.003\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.008\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTotal Effect\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.017\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e8.535\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.013\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.021\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.041\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eNote\u003c/strong\u003e: Bootstrap-based parameter estimates with 1,000 replications. Confidence intervals are based on the percentile method. All estimates are unstandardized unless noted. Std. Coef. Refers to standardized estimates (Std.all in lavaan). CESD10 = Depression score; SF-36 = Life satisfaction score; TPA = Total Physical Activity; Cognitive Impairment group: 0 = normal, 1 = impaired. ***p \u0026lt; 0.001; **p \u0026lt; 0.01; *p \u0026lt; 0.05.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 7. Standardized Regression Coefficients from SEM Model (Ordered by Magnitude)\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eDependent Variable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ePredictor\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eStd. Estimate\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ez-value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ep-value\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCI_group\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eEducation Level\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.302\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-72.215\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCESD10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eChronic Conditions\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.274\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e55.052\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCESD10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eEducation Level\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.156\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-34.023\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCI_group\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.148\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e30.059\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCI_group\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCESD10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.135\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e24.527\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLife Satisfaction (SF-36)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eChronic Conditions\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.119\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-21.796\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCESD10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSex\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-24.453\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLife Satisfaction (SF-36)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.092\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e17.096\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCESD10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMarital Status\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.081\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-15.920\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCESD10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTPA (z-score)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.069\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e14.344\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLife Satisfaction (SF-36)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMarital Status\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.053\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e9.679\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCI_group\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSex\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.044\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-9.806\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLife Satisfaction (SF-36)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTPA (z-score)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.042\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-8.370\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCI_group\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eChronic Conditions\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.039\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-8.077\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCI_group\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMarital Status\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.037\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-7.068\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCESD10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.033\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-6.733\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCI_group\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTPA (z-score)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.033\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e6.876\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCI_group\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLife Satisfaction (SF-36)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.027\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e5.287\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLife Satisfaction (SF-36)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSex\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.019\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e3.929\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLife Satisfaction (SF-36)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eEducation Level\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.007\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-1.403\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.161\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eNote\u003c/strong\u003e: Standardized regression coefficients (Std.all) from the SEM model are shown. TPA = Total Physical Activity (z-score); CESD10 = 10-item depression score; SF-36 = Life Satisfaction subscale; CI_group = Cognitive Impairment group (0 = normal, 1 = impaired). All estimates are based on 1,000 bootstrap replications. ***p \u0026lt; 0.001; **p \u0026lt; 0.01; *p \u0026lt; 0.05.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 8. Coefficient of Determination (R\u0026sup2;) for Endogenous Variables in the Structural Equation Model\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eEndogenous Variable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eR\u0026sup2;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInterpretation\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCESD10 (Depressive Symptoms)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.134\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e13.4% of the variance in depressive symptoms is explained by the predictors.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSF-36 Life Satisfaction\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.021\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.1% of the variance in life satisfaction is explained by the predictors.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCI_group (Cognitive Impairment Classification)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.174\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e17.4% of the variance in cognitive impairment status is explained by the model.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eNote\u003c/strong\u003e: R\u0026sup2; represents the proportion of variance in each endogenous variable explained by the model. Values were derived using the inspect(fit, \u0026quot;r2\u0026quot;) function from the lavaan package. CESD10 = 10-item Center for Epidemiologic Studies Depression Scale; SF-36 = Satisfaction With Life Scale; CI_group = Cognitive impairment group (0 = normal, 1 = impaired).\u003c/p\u003e\n\u003cp\u003eThe small positive direct path from TPA_z to cognitive impairment (\u0026beta; = 0.033) contrasts with the overall protective effect observed in the primary regression models. This likely reflects a statistical suppression effect, where the indirect protective mechanisms through improved mental health offset residual confounding or contextual strain associated with high physical activity levels. Readers are advised to interpret the SEM results in light of the nonlinear reverse J-shaped relationship identified in the primary analysis.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study provides robust evidence of a nonlinear, reverse J-shaped association between total physical activity (TPA) and cognitive impairment (CI) in a large, nationally representative sample of older Chinese adults. Using multivariable logistic regression and restricted cubic spline modeling, we identified an optimal activity threshold—approximately 2,800 MET-minutes per week—beyond which the cognitive benefits of physical activity appear to plateau or even decline. These results remained robust across multiple sensitivity models and stratified analyses. Furthermore, structural equation modeling revealed that depressive symptoms and life satisfaction partially mediated the TPA–cognition relationship, although the overall model exhibited complexity and limited global fit.\u003c/p\u003e\n\u003cp\u003eTogether, these findings suggest that while moderate physical activity plays a protective role in cognitive aging, excessively high activity volumes may not confer additional benefit—and may even be associa\u003c/p\u003e\n\u003cp\u003eted with subtle adverse effects. The complex interplay between behavioral, psychological, and physiological factors underscores the need for individualized activity recommendations tailored to demographic and health profiles. Below, we contextualize these findings within the existing literature, explore potential mechanisms, and discuss implications for intervention design and policy formulation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInterpretation of the Reverse J-Shaped Association and Underlying Mechanisms\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe reverse J-shaped association between total physical activity (TPA) and cognitive impairment (CI) suggests a nonlinear dose–response pattern in which moderate activity levels confer the most significant cognitive benefits, while insufficient and excessive activity may be suboptimal. Specifically, the lowest odds of cognitive impairment were observed at approximately 2,800 MET-minutes per week, as identified through restricted cubic spline modeling and confirmed in multivariable logistic regression models (Table 2, Figure 2). This threshold is notably higher than conventional public health guidelines, which typically recommend 500–1,000 MET-minutes per week \u003csup\u003e[27]\u003c/sup\u003e, and may reflect population-specific differences in baseline activity levels, occupation, and social norms surrounding physical exertion.\u003c/p\u003e\n\u003cp\u003eThis study identified a reverse J-shaped association between total physical activity (TPA) and cognitive impairment among older Chinese adults, with the lowest risk observed at approximately 2,800 MET-minutes per week. Compared to this optimal range, both insufficient and excessive levels of TPA were associated with increased odds of cognitive impairment. These findings suggest that the mental benefits of physical activity may plateau or even reverse beyond a certain threshold, highlighting the need for age-appropriate, moderate-intensity exercise guidelines in public health strategies.\u003c/p\u003e\n\u003cp\u003eNotably, individuals in the highest TPA quartile (≥5,000 MET-min/week) did not experience additional cognitive benefits and exhibited a modestly elevated risk of impairment (Table 2). This contrasts with the commonly assumed linear dose–response relationship and suggests a potential \"optimal dose window\" for cognitive protection. One possible explanation is that excessive physical activity—mainly when driven by occupational demands or low-autonomy contexts—may induce physiological stress, fatigue, or sleep disruption, which could offset its neuroprotective effects \u003csup\u003e[28- 30]\u003c/sup\u003e. While our study did not directly assess these mechanisms, the observed pattern warrants cautious interpretation and further investigation using longitudinal, mechanistic designs.\u003c/p\u003e\n\u003cp\u003eNotably, very high levels of total physical activity (particularly those exceeding 5,000 MET-min/week) were not associated with additional cognitive benefits and were instead linked to a modest increase in the risk of cognitive impairment (Table 2). This finding challenges the widely held assumption of a linear dose–response relationship and instead suggests an optimal physical activity range for maintaining cognitive health in older adults. Several potential explanations have been proposed in prior research. Studies—primarily based on animal models or limited human data—have indicated that excessive physical activity may induce adverse neurophysiological effects via mechanisms such as oxidative stress, chronic fatigue, or circadian rhythm disruption \u003csup\u003e[31,32]\u003c/sup\u003e. However, these mechanisms were not directly measured in the present analysis, and thus any causal interpretation should be made cautiously.\u003c/p\u003e\n\u003cp\u003eThese findings support the concept of an \"optimal dose window\" of physical activity for cognitive health in older adults, in which moderate activity provides a physiological stimulus sufficient to elicit neuroprotective adaptations. In contrast, extreme low or high activity levels fail to achieve—or may counteract—such effects. These insights offer a valuable refinement to existing linear-dose paradigms and suggest that public health recommendations may benefit from specifying upper and lower activity thresholds. In the following section, we further explore subgroup-specific differences in this dose–response relationship, particularly across sex, education level, and depressive symptomatology.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHigh-Volume Activity, Suppression Effects, and Model Fit Considerations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAlthough moderate physical activity levels consistently demonstrated protective effects against cognitive impairment, a notable nonlinear trend was observed. Specifically, participants in the highest quartile of total physical activity (Q4, ≥5,000 MET-min/week) showed a slightly elevated risk of cognitive impairment compared to those in the moderate activity group (Q2–Q3), deviating from the commonly assumed linear \"more is better\" hypothesis. This finding may be attributable to the type and context of physical activity among individuals with very high TPA—particularly when such activity is involuntary or occupational. Previous studies have suggested that these forms of physical exertion may lack psychological benefits, increase physiological strain, and impair recovery, thereby attenuating or reversing the cognitive benefits typically associated with physical activity \u003csup\u003e[31, 32, 36]\u003c/sup\u003e. Although this study did not stratify participants by activity type or socioeconomic status, the proposed explanation highlights the need for future research to account for the heterogeneity of activity contexts.\u003c/p\u003e\n\u003cp\u003eA second anomaly emerged in the structural equation modeling, where standardized TPA (TPA_z) showed a small but statistically significant positive direct association with cognitive impairment, even after adjusting for depressive symptoms and life satisfaction. This pattern contrasts with the overall protective effect observed in logistic regression and spline models, and likely reflects a statistical suppression effect. Specifically, the indirect protective pathways—through reduced depressive symptoms and enhanced life satisfaction—may be partially offset by residual confounding factors correlated with high TPA levels. For instance, older adults with high levels of obligatory physical activity (e.g., subsistence farming, caregiving) may experience cumulative fatigue, poor nutrition, or reduced social support, all contributing to cognitive decline. These complex psychosocial trade-offs are consistent with prior research suggesting that the mental benefits of physical activity are context-dependent and influenced by psychological and environmental mediators \u003csup\u003e[33]\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eFinally, we observed suboptimal global model fit in the SEM (RMSEA = 0.359, TLI \u0026lt; 0), suggesting potential structural misspecification or measurement limitations. Several factors may explain this outcome. First, using self-reported psychological mediators introduces shared method variance and imprecision in latent construct estimation. Second, the binary nature of the cognitive impairment outcome may reduce flexibility under maximum likelihood estimation. Third, modeling two parallel mediators—each reflecting distinct yet interrelated emotional constructs—may increase the risk of collinearity or overfitting, particularly in finite samples. Despite attempts to simplify the model (e.g., removing one mediator or reducing covariates), fit indices remained poor, suggesting that older adults' behavioral–cognitive relationship is inherently multifactorial. Future studies should consider employing latent moderated mediation, continuous cognitive outcomes, or accelerometer-based activity measures to improve explanatory power and model validity \u003csup\u003e[34, 35]\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eTaken together, these limitations suggest that the SEM findings should be interpreted with caution. Rather than serving as definitive evidence of mediation, the model offers a preliminary conceptual framework to illustrate possible psychosocial mechanisms underlying the reverse J-shaped association identified in the primary analysis. Future studies using longitudinal designs, objective PA measurement, and refined cognitive phenotyping are warranted to validate and expand upon these exploratory pathways.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePublic Health Implications and Tailored Interventions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIdentifying a reverse J-shaped association between total physical activity (TPA) and cognitive impairment (CI), with a turning point at approximately 2,800 MET-minutes/week, offers critical insight for public health recommendations. Although international guidelines often endorse a minimum of 500–1,000 MET-minutes/week for general health maintenance, our findings suggest that the cognitive benefits of physical activity may peak within a moderate range and diminish beyond a certain threshold. This nonlinearity necessitates the recalibration of physical activity targets in older populations, ensuring that intensity and volume are adjusted to avoid potential physiological or psychological overload. Elevated TPA may reflect health-adverse contexts rather than leisure-driven health behaviors, further complicating the dose-response relationship, particularly for individuals engaging in strenuous occupational or obligatory activity.\u003c/p\u003e\n\u003cp\u003ePopulation-level recommendations must also account for heterogeneous responses to physical activity. Our subgroup analyses revealed that individuals with lower education levels and women demonstrated heightened sensitivity to moderate TPA, potentially due to lower baseline cognitive reserve. This aligns with the cognitive reserve hypothesis, which posits that individuals with fewer educational or structural resources may benefit more from compensatory neuroprotective inputs such as physical activity \u003csup\u003e[36]\u003c/sup\u003e. For these subgroups, culturally and contextually appropriate interventions—such as tai chi, brisk walking, or group-based movement therapy—may be particularly effective. Simultaneously, strategies should address structural barriers to participation, especially in rural and underserved areas.\u003c/p\u003e\n\u003cp\u003eMoreover, the observed interaction between TPA and depressive symptoms suggests synergistic opportunities for mental and cognitive health promotion. Participants with higher baseline depression scores experienced greater cognitive benefits from moderate physical activity. This supports the integration of PA interventions within broader psychosocial and chronic disease frameworks, emphasizing programs that promote social engagement, emotional regulation, and health literacy \u003csup\u003e[37]\u003c/sup\u003e. Embedding physical activity initiatives into community-based services may enhance reach and sustainability, particularly for vulnerable older adults facing both mental health and cognitive risks.\u003c/p\u003e\n\u003cp\u003eTaken together, our findings reinforce the role of moderate physical activity as a modifiable, low-cost, and scalable non-pharmacological strategy for cognitive preservation in later life. However, activity recommendations must be tailored to individuals' occupational context, mental health status, and sociodemographic background to maximize benefit and minimize unintended consequences. Future policies should adopt a precision public health approach to promote cognitive resilience across diverse aging populations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStrengths and Limitations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study possesses several notable strengths. First, it draws upon a large, nationally representative dataset from the China Health and Retirement Longitudinal Study (CHARLS), enhancing the generalizability of our findings to older Chinese adults. The extensive sample size also enabled adequately powered stratified and interaction analyses across key subgroups, including sex, education, and vascular risk factors. Second, using a standardized cognitive composite score, based on validated multidomain assessments, provides a robust indicator of global cognitive functioning. Third, total physical activity (TPA) was quantified using the long-form International Physical Activity Questionnaire (IPAQ), a widely adopted tool with demonstrated reliability and cross-cultural applicability \u003csup\u003e[32,38,39]\u003c/sup\u003e. Although the IPAQ has limitations, its comprehensive assessment of activity domains (occupational, household, transportation, and leisure) is particularly relevant in capturing the diverse sources of energy expenditure among older adults in China. Fourth, we applied restricted cubic spline (RCS) modeling to flexibly characterize nonlinear dose–response relationships, avoiding oversimplified linear assumptions that may obscure essential threshold effects. Finally, extensive covariate adjustment—including demographics, chronic diseases, and depressive symptoms—enhanced the internal validity of our results, while complete-case analysis ensured consistency and reproducibility.\u003c/p\u003e\n\u003cp\u003eHowever, several limitations should be acknowledged. First, the cross-sectional design precludes causal inference. Although we observed a nonlinear association between TPA and cognitive impairment, the possibility of reverse causality—whereby cognitive decline reduces physical activity engagement—cannot be excluded. Moreover, given that mental decline is a progressive process, future studies should examine whether long-term or cumulative physical activity exposure better predicts cognitive trajectories, ideally through longitudinal cohort designs with repeated PA and cognition assessments. Longitudinal or interventional designs are needed to establish temporal directionality and evaluate whether increasing physical activity leads to sustained cognitive benefits. Second, TPA was self-reported via the IPAQ, making it vulnerable to recall inaccuracies and social desirability biases, particularly among cognitively impaired or older respondents \u003csup\u003e[36.38.39]\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eFurthermore, the IPAQ does not provide detailed data on activity intensity, duration per session, or domain-specific effects (e.g., leisure vs. occupational), limiting mechanistic interpretation. Third, cognitive impairment was defined using a single cutoff on a composite cognitive score, rather than clinically diagnosed dementia or mild cognitive impairment (MCI), potentially affecting diagnostic precision and cross-study comparability. Fourth, despite adjusting for key confounders, residual confounding from unmeasured variables—such as diet, genetic risk factors (e.g., APOE-ε4), air pollution, and social engagement—may still influence the observed associations \u003csup\u003e[31, 40]\u003c/sup\u003e. Fifth, we excluded individuals with missing data on key variables, possibly introducing selection bias. Although our complete-case approach enhanced data quality, future studies should consider multiple imputation or inverse probability weighting to improve robustness \u003csup\u003e[41, 42]\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eWhile this study provides novel insights into the nonlinear association between TPA and cognitive impairment in a nationally representative cohort of older Chinese adults, the findings should be interpreted cautiously. Future research employing prospective cohorts, accelerometer-based exposure assessment, and domain-specific cognitive measures will be critical to validate our conclusions and inform precise, population-tailored physical activity guidelines for cognitive aging.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eImplications and Future Directions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study offers novel insights into the dose–response relationship between total physical activity (TPA) and cognitive health in older Chinese adults. By identifying an inflection point near 2,800 MET-min/week, we contribute empirical evidence supporting a nonlinear, reverse J-shaped curve between TPA and cognitive impairment (CI). These findings challenge the dominant assumption of linear benefits from physical activity and suggest that moderate yet sustained physical activity may confer the most significant cognitive protection, while excessive levels may offer diminishing or harmful returns. This aligns with emerging neuroepidemiological perspectives emphasizing the balance between adaptive and maladaptive physiological responses to physical stress \u003csup\u003e[43-45]\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eFrom a public health standpoint, our results underscore the need to refine physical activity guidelines for aging populations in East Asia. Current WHO and national recommendations advocate 500–1,000 MET-min/week for general health benefits, but our findings indicate that older adults may benefit from aiming higher—approximately 2,000–3,000 MET-min/week—through accessible forms of aerobic activity such as walking, tai chi, or group calisthenics (see Figure 2). Notably, we caution against promoting overly intensive regimens, which may elicit adverse effects due to oxidative stress, inflammatory responses, or autonomic imbalance\u003csup\u003e\u0026nbsp;[29-31,46]\u003c/sup\u003e. These insights offer a basis for revising activity guidelines to reflect nonlinear dose–response patterns, especially in rapidly aging societies like China.\u003c/p\u003e\n\u003cp\u003eImportantly, personalization of interventions appears essential. Stratified analysis revealed that individuals with lower educational attainment and women derived greater cognitive benefit from physical activity, possibly due to lower baseline cognitive reserve and enhanced responsiveness to neuroprotective behaviors \u003csup\u003e[47-49]\u003c/sup\u003e. This suggests that education may serve as a key effect modifier, and tailoring interventions to vulnerable subgroups could maximize health gains while reducing inequality. The interaction effects observed reinforce the relevance of a \"precision prevention\" paradigm in cognitive aging research.\u003c/p\u003e\n\u003cp\u003eFuture studies should adopt longitudinal or experimental designs to validate the directionality and causality of the observed associations. Prospective cohorts with repeated TPA and cognition assessments and randomized controlled trials (RCTs) targeting varying activity doses are warranted to establish optimal thresholds. Additionally, integrating objective activity monitors (e.g., accelerometers, wearable sensors) may reduce recall bias inherent in self-report tools such as the IPAQ \u003csup\u003e[51-53]\u003c/sup\u003e. Mechanistic investigations using neuroimaging, inflammatory biomarkers, and sleep metrics would further illuminate the biological underpinnings of dose-specific effects.\u003c/p\u003e\n\u003cp\u003eTranslating these findings into real-world policy must account for structural barriers in rural and underserved areas. Community-based programs—such as walking groups, park fitness stations, or mobile health platforms—may provide scalable, low-cost strategies for promoting cognitive health in aging populations. These strategies are congruent with global aging policy frameworks, including the WHO's Global Action Plan on Physical Activity and China's Healthy Aging 2030 initiative, and warrant further implementation research \u003csup\u003e[25,53]\u003c/sup\u003e.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis nationally representative study observed a statistically significant reverse J-shaped association between total physical activity and cognitive impairment in older Chinese adults, with the lowest risk occurring around 2,800 MET-minutes per week. Both insufficient and excessive activity levels were associated with elevated cognitive risk. This pattern was partially mediated by depressive symptoms and life satisfaction, suggesting the importance of psychological well-being in this association. These findings refine the current understanding of the physical activity–cognition dose–response relationship and highlight the need for nuanced, context-aware activity guidelines. Public health strategies should prioritize personalized, culturally sensitive interventions—especially for individuals engaged in non-leisure physical labor, such as subsistence workers or manual caregivers. Future longitudinal and interventional studies using objective activity monitoring and detailed cognitive phenotyping are warranted to validate these thresholds and inform mental health promotion globally.\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe gratefully acknowledge all participants of the China Health and Retirement Longitudinal Study (CHARLS) and the CHARLS research team for providing access to the data.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eYongheng Zhao and Lizhen Ning contributed equally to the study design, data analysis, and manuscript drafting. Limeng Liu contributed to data interpretation and manuscript revision. Xuefeng Xi was responsible for statistical modeling and visualization. Gaixia Hou supervised the overall study and approved the final version of the manuscript. All authors have read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was supported by the Research Filing Project of the Department of Education of Heilongjiang Province, China (Project No. 1453ZD009). The funding agency was not involved in the study design, data collection, analysis, interpretation, or manuscript writing.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of Data and Materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe dataset analyzed in this study is publicly available from the China Health and Retirement Longitudinal Study (CHARLS) at http://charls.pku.edu.cn. Access to the data requires registration and permission from the CHARLS administrative team.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics Approval and Consent to Participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was conducted in accordance with the ethical principles of the Declaration of Helsinki (2013 revision) and was approved by the Institutional Review Board of Peking University (IRB00001052\u0026ndash;11015). Written informed consent was obtained from all participants prior to data collection.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical Trial Registration\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable. This study is based on secondary analysis of observational data from the publicly available CHARLS database.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for Publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eCorrigendum to: World Stroke Organization (WSO): Global Stroke Fact Sheet 2022. Int J Stroke. 2022;17(4):478. Epub 2022 Feb 22. Erratum for: Int J Stroke. 2022;17(1):18\u0026ndash;29.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eO'Donnell MJ, Chin SL, Rangarajan S, Xavier D, Liu L, Zhang H, Rao-Melacini P, Zhang X, Pais P, Agapay S, Lopez-Jaramillo P, Damasceno A, Langhorne P, McQueen MJ, Rosengren A, Dehghan M, Hankey GJ, Dans AL, Elsayed A, Avezum A, Mondo C, Diener HC, Ryglewicz D, Czlonkowska A, Pogosova N, Weimar C, Iqbal R, Diaz R, Yusoff K, Yusufali A, Oguz A, Wang X, Penaherrera E, Lanas F, Ogah OS, Ogunniyi A, Iversen HK, Malaga G, Rumboldt Z, Oveisgharan S, Al Hussain F, Magazi D, Nilanont Y, Ferguson J, Pare G, Yusuf S. INTERSTROKE investigators. Global and regional effects of potentially modifiable risk factors associated with acute stroke in 32 countries (INTERSTROKE): a case-control study. Lancet. 2016;388(10046):761\u0026ndash;75.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMeschia JF, Bushnell C, Boden-Albala B, Braun LT, Bravata DM, Chaturvedi S, Creager MA, Eckel RH, Elkind MS, Fornage M, Goldstein LB, Greenberg SM, Horvath SE, Iadecola C, Jauch EC, Moore WS, Wilson JA, American Heart Association Stroke Council; Council on Cardiovascular and Stroke Nursing. Council on Clinical Cardiology; Council on Functional Genomics and Translational Biology; Council on Hypertension. Guidelines for the primary prevention of stroke: a statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2014;45(12):3754\u0026ndash;832.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePowell-Wiley TM, Poirier P, Burke LE, Despr\u0026eacute;s JP, Gordon-Larsen P, Lavie CJ, Lear SA, Ndumele CE, Neeland IJ, Sanders P, St-Onge MP, American Heart Association Council on Lifestyle and Cardiometabolic Health; Council on Cardiovascular and Stroke Nursing. Council on Clinical Cardiology; Council on Epidemiology and Prevention; and Stroke Council. Obesity and Cardiovascular Disease: A Scientific Statement From the American Heart Association. Circulation. 2021;143(21):e984\u0026ndash;1010.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGuthold R, Stevens GA, Riley LM, Bull FC. Worldwide trends in insufficient physical activity from 2001 to 2016: a pooled analysis of 358 population-based surveys with 1\u0026middot;9 million participants. Lancet Global Health. 2020;8(1):e39\u0026ndash;51.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCampbell BCV, Khatri P, Stroke. Lancet. 2020;396(10244):129\u0026ndash;42.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSankaran VG, Weiss MJ. Anemia: progress in molecular mechanisms and therapies. Nat Med. 2015;21(3):221\u0026ndash;30.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTian W, Zhu G, Xiao W, Gao B, Lu W, Wang Y. Stroke burden and attributable risk factors in China, 1990\u0026ndash;2019. Front Neurol. 2023;14:1193056.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMa Q, Li R, Wang L, Yin P, Wang Y, Yan C, Ren Y, Qian Z, Vaughn MG, McMillin SE, Hay SI, Naghavi M, Cai M, Wang C, Zhang Z, Zhou M, Lin H, Yang Y. Temporal trend and attributable risk factors of stroke burden in China, 1990\u0026ndash;2019: an analysis for the Global Burden of Disease Study 2019. Lancet Public Health. 2021;6(12):e897\u0026ndash;906.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGBD 2019 Diseases and Injuries Collaborators. Global burden of 369 diseases and injuries in 204 countries and territories, 1990\u0026ndash;2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet. 2020;396(10258):1204\u0026ndash;22.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu X, Zhang D, Liu Y, Sun X, Han C, Wang B, Ren Y, Zhou J, Zhao Y, Shi Y, Hu D, Zhang M. Dose-Response Association Between Physical Activity and Incident Hypertension: A Systematic Review and Meta-Analysis of Cohort Studies. Hypertension. 2017;69(5):813\u0026ndash;20.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLaMonte MJ, Blair SN, Church TS. Physical activity and diabetes prevention. J Appl Physiol. 2005;99(3):1205\u0026ndash;13.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLubogo D, Wamani H, Mayega RW, Orach CG. Effects of nutrition education, physical activity and motivational interviewing interventions on metabolic syndrome among females of reproductive age in Wakiso district, central Uganda: a randomised parallel-group trial. BMC Public Health. 2025;25(1):790.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eErickson KI, Voss MW, Prakash RS, Basak C, Szabo A, Chaddock L, Kim JS, Heo S, Alves H, White SM, Wojcicki TR, Mailey E, Vieira VJ, Martin SA, Pence BD, Woods JA, McAuley E, Kramer AF. Exercise training increases size of hippocampus and improves memory. Proc Natl Acad Sci U S A. 2011;108(7):3017\u0026ndash;22.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHosseininasabnajar F, Kakinami L, Alzheimer\u0026rsquo;s Disease Neuroimaging Initiative. The longitudinal effects of global and regional brain measurements on cognitive abilities. Sci Rep. 2025;15(1):27846.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBenedict C, Cedernaes J. Reflections on The Lancet's Commission on dementia prevention, intervention, and care. Lancet. 2025;405(10479):624\u0026ndash;5.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eso-Markku P, Aaltonen S, Kujala UM, Halme HL, Phipps D, Knittle K, et al. Physical activity and cognitive decline among older adults: a systematic review and meta-analysis. JAMA Netw Open. 2024;7(2):e2354285.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBooth FW, Roberts CK, Laye MJ. Lack of exercise is a major cause of chronic diseases. Compr Physiol. 2012;2(2):1143\u0026ndash;211.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGhozy S, Zayan AH, El-Qushayri AE, Parker KE, Varney J, Kallmes KM, Morsy S, Abbas AS, Diestro JDB, Dmytriw AA, Shah J, Hassan AE, Islam SMS. Physical activity level and stroke risk in US population: A matched case-control study of 102,578 individuals. Ann Clin Transl Neurol. 2022;9(3):264\u0026ndash;75.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFan M, Lyu J, He P. Zhonghua Liu Xing Bing Xue Za Zhi. 2014;35(8):961\u0026ndash;4. Chinese. PMID: 25376692. [Chinese guidelines for data processing and analysis concerning the International Physical Activity Questionnaire].\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhao Y, Hu Y, Smith JP, Strauss J, Yang G. Cohort profile: the China Health and Retirement Longitudinal Study (CHARLS). Int J Epidemiol. 2014;43(1):61\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhao Y, Strauss J, Yang G, Giles J, Hu P, Hu Y, et al. China Health and Retirement Longitudinal Study (CHARLS) 2011\u0026ndash;2012 national baseline users\u0026rsquo; guide. Beijing: National School of Development, Peking University; 2013. pp. 1\u0026ndash;56.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eIPAQ Research Committee. Guidelines for data processing and analysis of the International Physical Activity Questionnaire (IPAQ) \u0026ndash; Short and Long Forms [Internet]. Revised November 2005 [cited 2025 Jul 31]. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://sites.google.com/view/ipaq/score\u003c/span\u003e\u003cspan address=\"https://sites.google.com/view/ipaq/score\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWu YT, Beiser AS, Breteler MMB, Fratiglioni L, Helmer C, Hendrie HC, Honda H, Ikram MA, Langa KM, Lobo A, Matthews FE, Ohara T, P\u0026eacute;r\u0026egrave;s K, Qiu C, Seshadri S, Sj\u0026ouml;lund BM, Skoog I, Brayne C. The changing prevalence and incidence of dementia over time - current evidence. Nat Rev Neurol. 2017;13(6):327\u0026ndash;39.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLei X, Hu Y, McArdle JJ, Smith JP, Zhao Y. Gender differences in cognition among older adults in China. J Hum Resour. 2012;47(4):951\u0026ndash;71. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3368/jhr.47.4.951\u003c/span\u003e\u003cspan address=\"10.3368/jhr.47.4.951\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAndresen EM, Malmgren JA, Carter WB, Patrick DL. Screening for depression in well older adults: evaluation of a short form of the CES-D (CESD-10). Am J Prev Med. 1994;10(2):77\u0026ndash;84.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWorld Health Organization. Global Recommendations on Physical Activity for Health. Geneva: WHO. 2010. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.who.int/publications/i/item/9789241599979\u003c/span\u003e\u003cspan address=\"https://www.who.int/publications/i/item/9789241599979\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMaeda S, Tanabe T, Otsuki T, Sugawara J, Iemitsu M, Miyauchi T, Kuno S, Ajisaka R, Matsuda M. Moderate regular exercise increases basal production of nitric oxide in elderly women. Hypertens Res. 2004;27(12):947\u0026ndash;53.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFerrer MD, Cap\u0026oacute; X, Martorell M, Busquets-Cort\u0026eacute;s C, Bouzas C, Carreres S, Mateos D, Sureda A, Tur JA, Pons A. Regular Practice of Moderate Physical Activity by Older Adults Ameliorates Their Anti-Inflammatory Status. Nutrients. 2018;10(11):1780.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOtocka-Kmiecik A, Orłowska-Majdak M, Stawski R, Szkudlarek U, Kosielski P, Padula G, Gałczyński S, Nowak D. Repetitions of Strenuous Exercise Consistently Increase Paraoxonase 1 Concentration and Activity in Plasma of Average-Trained Men. Oxid Med Cell Longev. 2021;2021:2775025.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eManolis AS, Manolis AA. Exercise and Arrhythmias: A Double-Edged Sword. Pacing Clin Electrophysiol. 2016;39(7):748\u0026ndash;62.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZotcheva E, Bratsberg B, Strand BH, Jugessur A, Engdahl BL, Bowen C, Selb\u0026aelig;k G, Kohler HP, Harris JR, Weiss J, Tom SE, Krokstad S, Mekonnen T, Edwin TH, Stern Y, H\u0026aring;berg AK, Skirbekk V. Trajectories of occupational physical activity and risk of later-life mild cognitive impairment and dementia: the HUNT4 70\u0026thinsp;+\u0026thinsp;study. Lancet Reg Health Eur. 2023;34:100721.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHa JC, Yoon JH. Differential associations of physical job demands with cognitive impairment in Korean workers aged 45 and older: a 14-year longitudinal study using the Korean Longitudinal Study of Aging (KLoSA). BMJ Open. 2024;14(10):e088000.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRubio Casta\u0026ntilde;eda FJ, Tom\u0026aacute;s Aznar C, Muro Baquero C. Medici\u0026oacute;n de la actividad f\u0026iacute;sica en personas mayores de 65 a\u0026ntilde;os mediante el IPAQ-E: validez de contenido, fiabilidad y factores asociados [Validity, Reliability and Associated Factors of the International Physical Activity Questionnaire Adapted to Elderly (IPAQ-E)]. Rev Esp Salud Publica. 2017;91:e201701004. Spanish. PMID: 28098134.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZheng C, Gill JMR, Sun FH, Huang WY, Sheridan S, Chen XK, Wu Y, Wong CK, Tian XY, Wong SH. Effects of increasing light versus moderate-to-vigorous physical activity on cardiometabolic health in Chinese adults with obesity. J Sports Sci. 2023;41(16):1547\u0026ndash;57.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBarbosa Dos Santos R, Kaur N, Nelson M, Catizzone M, Sheehy L, Munce S, Inness EL, Salbach NM. Feasibility, safety, and potential benefit of a virtual, community-based, task-oriented exercise program (TIMETM at Home) for people with balance and mobility limitations: a pre-post feasibility study. Disabil Rehabil 2025 Jul 31:1\u0026ndash;20.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJohnsen B, Strand BH, Martinaityte I, Lorem GF, Schirmer H. Leisure Time Physical Activities' Association With Cognition and Dementia: A 19 Years' Life Course Study. Front Aging Neurosci. 2022;14:906678.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003e\u0026Ouml;hlin J, Toots A, Dahlin Almevall A, Littbrand H, Conradsson M, H\u0026ouml;rnsten C, Werneke U, Niklasson J, Olofsson B, Gustafson Y, Wennberg P, S\u0026ouml;derberg S. Concurrent validity of the International Physical Activity Questionnaire adapted for adults aged\u0026thinsp;\u0026ge;\u0026thinsp;80 years (IPAQ-E 80 +) - tested with accelerometer data from the SilverMONICA study. Gait Posture. 2022;92:135\u0026ndash;43.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNoguchi KS, Sansanwal S, Mehdipour A, Tang A. Comparing the reliability of physical activity questionnaires in community-dwelling adults with stroke. Top Stroke Rehabil. 2025;32(2):130\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBrenner PS, DeLamater JD. Social desirability bias in self-reports of physical activity: is an exercise identity the culprit? Soc Indic Res. 2014;117(2):489\u0026ndash;504.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eValenzuela PL, Ruilope LM, Santos-Lozano A, Wilhelm M, Kr\u0026auml;nkel N, Fiuza-Luces C, Lucia A. Exercise benefits in cardiovascular diseases: from mechanisms to clinical implementation. Eur Heart J. 2023;44(21):1874\u0026ndash;89.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAustin PC, White IR, Lee DS, van Buuren S. Missing Data in Clinical Research: A Tutorial on Multiple Imputation. Can J Cardiol. 2021;37(9):1322\u0026ndash;31.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMukaka M, White SA, Terlouw DJ, Mwapasa V, Kalilani-Phiri L, Faragher EB. Is using multiple imputation better than complete case analysis for estimating a prevalence (risk) difference in randomized controlled trials when binary outcome observations are missing? Trials. 2016;17:341.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHu Y, Peng W, Ren R, Wang Y, Wang G. Sarcopenia and mild cognitive impairment among elderly adults: The first longitudinal evidence from CHARLS. J Cachexia Sarcopenia Muscle. 2022;13(6):2944\u0026ndash;52.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eQin F, Luo M, Xiong Y, Zhang N, Dai Y, Kuang W, Cen X. Prevalence and associated factors of cognitive impairment among the elderly population: A nationwide cross-sectional study in China. Front Public Health. 2022;10:1032666.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLee CD, Folsom AR, Blair SN. Physical activity and stroke risk: a meta-analysis. Stroke. 2003;34(10):2475\u0026ndash;81.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFenyo IM, Gafencu AV. The involvement of the monocytes/macrophages in chronic inflammation associated with atherosclerosis. Immunobiology. 2013;218(11):1376\u0026ndash;84.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSun D, Liu C, Ding Y, Yu C, Guo Y, Sun D, Pang Y, Pei P, Du H, Yang L, Chen Y, Meng X, Liu Y, Liu J, Sohoni R, Sansome G, Chen J, Chen Z, Lv J, Kan H, Li L. China Kadoorie Biobank Collaborative Group. Long-term exposure to ambient PM2\u0026middot;5, active commuting, and farming activity and cardiovascular disease risk in adults in China: a prospective cohort study. Lancet Planet Health. 2023;7(4):e304\u0026ndash;12.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi J, Zhang X, Zhang M, Wang L, Yin P, Li C, You J, Huang Z, Ng M, Wang L, Zhou M. Urban-rural differences in the association between occupational physical activity and mortality in Chinese working population: evidence from a nationwide cohort study. Lancet Reg Health West Pac. 2024;46:101083.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCarroll S, Dudfield M. What is the relationship between exercise and metabolic abnormalities? A review of the metabolic syndrome. Sports Med. 2004;34(6):371\u0026ndash;418.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBrenner P, Delamater J. Social desirability bias in self-reports of physical activity: Is an exercise identity the culprit? Soc Indic Res. 2014;117:456.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCleland C, Ferguson S, Ellis G, Hunter RF. Validity of the International Physical Activity Questionnaire (IPAQ) for assessing moderate-to-vigorous physical activity and sedentary behaviour of older adults in the United Kingdom. BMC Med Res Methodol. 2018;18(1):176.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFini NA, Simpson D, Moore SA, Mahendran N, Eng JJ, Borschmann K, Moulaee Conradsson D, Chastin S, Churilov L, English C. How should we measure physical activity after stroke? An international consensus. Int J Stroke. 2023;18(9):1132\u0026ndash;42.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDiep L, Kwagyan J, Kurantsin-Mills J, Weir R, Jayam-Trouth A. Association of physical activity level and stroke outcomes in men and women: a meta-analysis. J Womens Health (Larchmt). 2010;19(10):1815\u0026ndash;22.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Cognitive impairment, Physical activity, Dose–response relationship, Older adults, CHARLS","lastPublishedDoi":"10.21203/rs.3.rs-7282499/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7282499/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground:\u003c/strong\u003e Cognitive impairment (CI) is a growing concern in aging societies, particularly in low- and middle-income countries. Although physical activity (PA) is widely recognized as neuroprotective, its optimal dose for cognitive health remains uncertain.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eObjective\u003c/strong\u003e: To evaluate the nonlinear association between total physical activity (TPA) and cognitive impairment in older Chinese adults, and to examine whether depressive symptoms and life satisfaction mediate this relationship.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods:\u003c/strong\u003e We analyzed data from 7,818 adults aged ≥60 years in the China Health and Retirement Longitudinal Study (CHARLS, 2011–2020). Total physical activity (TPA) was assessed using the long-form IPAQ and expressed in MET-minutes/week. CI was defined as a total cognitive score \u0026lt;10. Logistic regression, restricted cubic spline models, and structural equation modeling (SEM) were used to examine nonlinear associations and psychological mediation pathways, adjusting for demographic and health covariates.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults:\u003c/strong\u003e We observed a reverse J-shaped association between TPA and CI risk, with the lowest odds at ~2,800 MET-min/week (OR = 0.772, 95% CI: 0.648–0.917). Excessive TPA was not associated with additional benefits and showed a trend toward increased risk. Mediation analysis revealed that depressive symptoms and life satisfaction partially accounted for the relationship.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion:\u003c/strong\u003e This study observed a statistically significant reverse J-shaped association between total physical activity and cognitive impairment in older Chinese adults, with the lowest risk occurring around 2,800 MET-minutes per week. Both insufficient and excessive activity levels were associated with higher cognitive risk, extending previous findings by quantifying this population's optimal activity range for cognitive health. Additionally, depressive symptoms and life satisfaction were identified as partial mediators, suggesting that psychological well-being may play a significant role in the pathway linking physical activity to cognitive outcomes among older adults.\u003c/p\u003e","manuscriptTitle":"A Reverse J-Shaped Association Between Total Physical Activity and Cognitive Impairment in Older Chinese Adults: Evidence from a Nationally Representative Cross-sectional Study Using CHARLS Data","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-26 14:21:16","doi":"10.21203/rs.3.rs-7282499/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"cae3b5a5-59f8-4438-b57f-fa6ffe836a77","owner":[],"postedDate":"August 26th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-09-26T05:54:05+00:00","versionOfRecord":[],"versionCreatedAt":"2025-08-26 14:21:16","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7282499","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7282499","identity":"rs-7282499","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}