Abstract
To determine the dose-response association between physical activity (PA) and all-cause mortality to identify minimum PA thresholds for significant mortality risk reduction by using a nationally representative sample of Chinese adults. This population-based cohort study included 109,407 Chinese participants from the Kailuan study with linkage to national death records up to December 31, 2022. Baseline data were derived from self-reported surveys among participants aged 53.32 ± 13.55 years. During a median follow-up of 6.99 (IQR 5.99–6.99) years, 4571 participants died from all causes. Performing at least 75 min per week of physical activity contributed to significantly reduced all-cause mortality. Compared with physical inactivity, total physical activity was inversely associated with the risk of mortality in the very low (HR: 1.08; 95% CI: 0.98–1.19), low (HR: 0.89; 95% CI: 0.81–0.98), medium (HR: 0.88; 95% CI: 0.79–0.96) and high (HR: 0.77; 95% CI: 0.69–0.85) quartiles. The dose‒response curve revealed a steady decline in HRs with more minutes of physical activity per week with all-cause mortality. Our study demonstrated an inverse nonlinear dose‒response relationship between physical activity and all-cause mortality in elderly adults, with as little as 75 min of weekly activity showing clinically meaningful risk reduction compared with inactivity. These findings suggest that mortality benefits at lower thresholds for elderly adults. The results provide evidence-based justification for health policies that promote graduated physical activity targets to improve population-level adherence among aging communities.
Supplementary Information
The online version contains supplementary material available at 10.1038/s41598-025-26356-8.
Keywords: Physical activity, Mortality, Vigorous physical activity, Moderate physical activity, All-cause mortality, Dose‒response relationship, Lifestyle, Life expectancy
Subject terms: Diseases, Health care, Risk factors
Introduction
Physical activity (PA) has been recommended by the World Health Organization (WHO) for its significant health benefits for hearts, bodies, and minds in their physical activity guidelines1. The beneficial mechanisms of physical activity have been explained by a wide range of studies on preventing and managing noncommunicable diseases (NCDs), such as heart disease, stroke, diabetes, and several cancers2,3. Physical activity is also effective in reducing adiposity4, improving sleep quality5, and reducing symptoms of anxiety and depression6. Additionally, evidence from observational studies has demonstrated significant associations of physical activity with lower risks of subsequent mortality7.
The WHO’s 2020 guidelines on physical activity indicated a lack of precise dose‒response data between PA and mortality8. Although systematic reviews and meta-analyses have provided strong evidence that both low and high amounts of PA could reduce the risk of mortality in adults9,10, an existing study from a Japanese cohort revealed a nonlinear dose–response relationship11. Regarding the health benefits of accumulated physical activity, studies exploring the dose‒response relationship between PA minutes and mortality in China is necessary. Only one study examined the PA dose-dependent relationship with pneumonia-specific mortality among Chinese adults12. Unlike revealing a dose‒response relationship between PA and all-cause mortality in populations with chronic diseases13, our analysis excluded individuals with major comorbidities to specifically assess this association in healthier elderly adults. Considering the inevitable mortality risk associated with chronic diseases, it is essential to identify the dose‒response curve of PA in relation to all-cause mortality among the general population.
Despite established links between physical activity and mortality, evidence remains scarce for the Chinese adult population, particularly regarding precise dose‒response thresholds. To address this gap, this prospective cohort study aims to (1) quantify the nonlinear association between PA volume and all-cause mortality in a comorbidity-free Chinese population and (2) identify minimum PA thresholds for significant mortality risk reduction. By excluding major chronic diseases and employing rigorous covariate adjustment, our study provides tailored recommendations for healthy aging in China’s rapidly growing elderly population.
Methods
Study participants
This study utilized data from the Kailuan Study, an ongoing healthcare examination that includes surveys and health examinations conducted every 2 years since 2006. The study is funded by Tangshan Kailuan Ltd. in China. Detailed information about the Kailuan study design, survey methods, clinical examinations and participant characteristics has been described elsewhere14,15. Potential participants were selected via a multistage design and included a total of 109,407 adult participants selected from the years 2014 and 2016. Baseline information, including lifestyle behaviors, social status, health status, and disease history, was obtained from each registered participant.
These baseline data were collected by trained doctors and nurses from 11 hospitals within the KaiLuan Group, who are responsible for healthcare. It was linked to clinically registered death records from Kailuan hospitals up to December 31, 2022. Participants with missing responses regarding physical activity habits were excluded (n = 13315). Additionally, participants with missing survey time data (n = 131) and/or missing death records (n = 893) were also excluded. A total of 109,407 participants were included in the final statistical analysis (Fig. 1). Similar characteristics were found between the exclude and non-exclude participants (eTable 6). The study was approved by the KaiLuan General Hospital Ethics Committee, and written informed consent was obtained from all participants. The study was performed in accordance with relevant guidelines.
Fig. 1.
Flow diagram of study participants. Abbreviations: VPA vigorous physical activity; MPA moderate physical activity; WPA walking physical activity.
Assessment of physical activity
Physical activity was recorded by using the standardized International Physical Activity Questionnaire (IPAQ) short form. During the baseline survey, participants were asked about the frequency and duration of physical activities at different intensity levels (including vigorous physical activity, moderate physical activity and walking activity) during the past week. The frequency of physical activity was reported as days per week (“<1 day”, “1–2 days”, “3–4 days”, “5–6 days”, “every day”)16. The duration of physical activity was recorded in minutes of exercise (“Never”, “<30 minutes”, “30–60 minutes”, “>60 minutes”). Total physical activity (TPA) was calculated as the sum of weighted vigorous physical activity, weighted moderate physical activity and weighted walking activity by multiplying physical activity frequency and duration by weight (frequency * duration*60, frequency * duration*45, frequency * duration*15). The formula referred to an existing study using formula : “weekly total MET minutes = weekly vigorous exercise days × daily vigorous exercise duration (minutes) × (8 METs) + weekly moderate exercise days × daily moderate exercise duration (minutes) × (4 METs) + weekly walk exercise days × daily walk exercise duration (minutes) × (3.3 METs)”17.
Assessment of covariates
Covariates were assessed at baseline and analyzed as time-invariant variables. Biennial demographic data, including age, sex, level of education, level of income, marital status, body height and body weight, lifestyle factors, such as smoking habits, drinking habits and use of screen devices, work-related data, such as work type and intensity, and disease history including Parkinson’s disease (PD), stroke, fatty liver, heart failure, and myocardial infarction, were collected through standardized questionnaires and clinical examinations at baseline and follow-up. For level of education, we categorized education into three levels: “illiteracy/primary school”, “middle/high school”, and “college or above”. Regarding income, we grouped participants into “<1000yuan”, “1000-3000yuan”, and “>=3000yuan”. Body height and weight were measured by trained staff using well-calibrated instruments, and body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared. Smoking habits were assessed by asking about the number of cigarettes consumed per day, categorized as “<= 10 cigarettes”, “11–20 cigarettes”, or “>=20 cigarettes”. Drinking habits were divided into non-drinkers and drinkers. The participants self-reported their work type and occupational physical intensity, with work intensity classified into four categories: “no intensity”, “light intensity”, “moderate intensity”, and “heavy intensity”. Missing covariate data were handled through single imputation: continuous variables were replaced with the mean value, and categorical variables were replaced with the mode (dominant category).
Ascertainment of mortality
All participants were followed up every two years until death or December 31, 2022. Information regarding the cause of death was obtained from Kailuan Hospital records, death certificates from provincial vital statistics offices, and medical records from medical insurance or hospitals. Death ascertains the required confirmation from ≥ 2 sources (e.g., hospital + insurance records). Discordant cases were resolved by prioritizing provincial certificates and then physician review by using standardized diagnostic criteria (i.e., ICD-10 codes).
Statistical analysis
For baseline characteristics, continuous variables are described as medians or means with standard deviations (SEs), and categorical variables are presented as proportions (%). Comparisons of continuous or categorical variables were conducted via one-way ANOVA tests or Chi-square tests.
By categorizing participants into physically inactive and physically active groups, the physically active participants were further divided into quartiles on the basis of total physical activity volume per week (“<75 min/wk”, “75–360 min/wk”, “360–720 min/wk”, “>720 min/wk”), and characteristics were compared across these groups. Person-years were calculated from the baseline examination date to the date of death or the end of follow-up (December 31, 2022), whichever came first. We then analyzed the associations between minutes of physical activity and all-cause mortality via Cox proportional hazards models. In the initial model (Model 1), hazard ratios (HRs) and 95% confidence intervals (CIs) of mortality were estimated, adjusting for age and sex. Model 2 included additional adjustments for potential confounders such as cigarette intake, drinking habits, screen time, marital status, educational level and income level. Model 3 was further adjusted for BMI, history of Parkinson’s disease (PD), stroke, heart failure, fatty liver and myocardial infarction. Model 4 added adjustments for work intensity and work type in addition to the variables in Model 3.
To investigate the dose‒response relationship between total physical activity and the risk of all-cause mortality, we utilized a restricted cubic spline (RCS) regression with 4 knots placed at the 5th, 35th, 65th, and 95th percentiles. We employed the Akaike information criterion (AIC) to select the restricted cubic spline with 4 knots and chose the model with the lowest AIC. We designated the physically inactive group with 0 min/week of TPA as the reference group.
Subgroup analysis was performed using a forest plot. Comparisons were made between participants below and above 65 years of age, male and female, brain workers and manual workers, participants in active positions or retired, and participants with a medical history of stroke, fatty liver, or heart failure. We further examined the combined associations of age and physical activity with all-cause mortality. A heatmap was created to display the joint effects as age increased every 10 years from 18 to 99 years.
Finally, several sensitivity analyses were performed: (1) by excluding participants who died in the first two years of follow-up and removing participants with less than five years of follow-up and (2) by excluding participants with a history of stroke, heart failure or fatty liver disease. Furthermore, we examined the correlations between vigorous physical activity, moderate physical activity, and walking activity. We also investigated the potential confounding effect of BMI on physical activity for all-cause mortality. Statistical analyses were conducted using SAS (version 9.4) and R (version 4.2.3). A two-sided P < 0.05 was used to determine statistical significance.
Results
Population characteristics
In this study, 109,407 eligible adults were included, among whom 78.9% were men. At baseline, the mean age of the included participants was 53.3 (SD = 13.6) years, and the mean BMI was 24.9 (SD = 3.38) kg/m2. The education level and average monthly income for Chinese people were relatively low, with 85.85% having a middle or high school level of education and 73.45% earning less than 1000 Chinese yuan. The majority of participants were manual workers (89.80%), whereas only 10.20% were office workers. A total of 48.66% of people had no physical activity habits, and for those with physical activity habits, the mean duration was 224.63 min/wk. For those who were physically active, the TPA was categorized into quartiles (“0–75 min/wk”, “75–360 min/wk”, “360–720 min/wk”, “>720 min/wk”). Table 1 provides more details by quartiles of TPA durations per week.
Table 1.
Baseline characteristics of participants by total physical activity per week (N = 109407).
| Characteristics | Total Physical Activity per Week, Mean (SD) | Overall | ||||
|---|---|---|---|---|---|---|
| No TPA | Quartile 1 (0 < TPA < 75 min) |
Quartile 2 (75 < = TPA < 360 min) |
Quartile 3 (360 < = TPA < 720 min) |
Quartile 4 (TPA > 720 min) |
||
| Total participants No. (%) | 53,240(48.66) | 11,973(10.94) | 15,043(13.75) | 14,366(13.13) | 14,785(13.51) | 109,407(100) |
| Men No. (%) | 40,471(76.02) | 9743(81.37) | 12,299(81.76) | 11,626(80.93) | 12,228(82.71) | 86,367(78.94) |
| Age, year | 54.54 ± 13.15 | 52.86 ± 14.23 | 51.87 ± 13.69 | 52.40 ± 13.51 | 51.69 ± 13.89 | 53.32 ± 13.55 |
| BMI | 24.93 ± 3.38 | 24.89 ± 3.37 | 24.90 ± 3.41 | 24.89 ± 3.31 | 24.93 ± 3.37 | 24.92 ± 3.38 |
| Marriage status No. (%) | ||||||
| Single | 834(1.57) | 438(3.66) | 477(3.17) | 438(3.05) | 503(3.40) | 2263(2.07) |
| Married | 51,571(96.87) | 11,253(93.99) | 14,262(94.81) | 13,568(94.45) | 13,863(93.76) | 104,965(95.94) |
| Divorced | 376(0.71) | 113(0.94) | 121(0.80) | 138(0.96) | 155(1.05) | 873(0.80) |
| Widowed | 314(0.59) | 119(0.99) | 118(0.78) | 153(1.07) | 178(1.20) | 920(0.84) |
| Re-married | 145(0.27) | 50(0.42) | 65(0.43) | 69(0.48) | 86(0.58) | 384(0.35) |
| Education level No. (%) | ||||||
| Illiteracy/primary school | 688(1.29) | 798(6.66) | 612(4.07) | 397(2.76) | 245(1.66) | 2740(2.50) |
| Middle/high school | 48,659(91.40) | 9160(76.51) | 12,002(79.78) | 11,342(78.95) | 12,758(86.29) | 93,919(85.85) |
| College or above | 3893(7.31) | 2015(16.83) | 2429(16.15) | 2627(18.29) | 1782(12.05) | 12,746(11.65) |
| Income status No. (%) | ||||||
| <1000 Chinese yuan | 44,744(84.04) | 7422(61.99) | 9710(64.55) | 8842(61.55) | 9638(65.19) | 80,354(73.45) |
| 1000–3000 Chinese yuan | 6193(11.63) | 3739(31.23) | 4156(27.63) | 3899(27.14) | 2797(18.92) | 20,784(19.00) |
| >=3000 Chinese yuan | 2303(4.33) | 812(6.78) | 1177(7.82) | 1625(11.31) | 2350(15.89) | 8267(7.56) |
| Work type No. (%) | ||||||
| Brain worker | 4045(7.60) | 1635(13.66) | 1876(12.47) | 1977(13.76) | 1676(11.34) | 11,156(10.20) |
| Manual worker | 49,195(92.40) | 10,338(86.34) | 13,167(87.53) | 12,389(86.24) | 13,109(88.66) | 98,249(89.80) |
| Cigarette intake No. (%) | ||||||
| <= 10 cigarettes | 52,252(98.14) | 11,243(93.90) | 13,960(92.80) | 13,539(94.24) | 13,566(91.76) | 99,099(90.58) |
| 11–20 cigarettes | 819(1.54) | 661(5.52) | 957(6.36) | 738(5.14) | 998(6.75) | 8974(8.20) |
| >= 20 cigarettes | 169(0.32) | 69(0.58) | 126(0.84) | 89(0.62) | 221(1.49) | 1332(1.22) |
| Drinking habits No. (%) | ||||||
| No drinking | 34,612(65.01) | 8953(74.78) | 11,455(76.15) | 11,296(78.63) | 9032(61.09) | 75,347(68.87) |
| Drinker | 18,628(34.99) | 3020(25.22) | 3588(23.85) | 3070(21.37) | 5752(38.91) | 34,058(31.13) |
| Work intensity No. (%) | ||||||
| No intensity | 39,582(74.35) | 6583(54.98) | 9490(63.09) | 8755(60.94) | 9446(63.89) | 73,855(67.51) |
| Light intensity | 5696(10.70) | 3019(25.22) | 2730(18.15) | 2817(19.61) | 3176(21.48) | 17,437(15.94) |
| Moderate intensity | 5234(9.83) | 1230(10.27) | 1588(10.56) | 1601(11.14) | 1431(9.68) | 11,084(10.13) |
| Heavy intensity | 2728(5.12) | 1141(9.53) | 1235(8.21) | 1193(8.30) | 732(4.95) | 7029(6.42) |
| Use of screen devices (hours) | 2.35 ± 1.58 | 2.41 ± 1.54 | 2.33 ± 1.53 | 2.39 ± 1.60 | 2.48 ± 1.57 | 2.37 ± 1.57 |
Abbreviation: TPA Total physical activity; SD standard deviation.
Variables are displayed in percentage or mean ± SD.
Association of total physical activity duration with mortality
During a median follow-up of 6.99 (IQR 5.99–6.99) years, 4571 participants died from all causes. Compared with those in the physically inactive group, those who engaged in physical activity for at least 75 min per week had significantly lower all-cause mortality, as indicated by adjusted hazard ratios (HRs) and 95% confidence intervals (CIs). Total physical activity showed an inverse relationship with the risk of mortality, with HRs of 1.08 (95% CI: 0.98–1.19) for the very low quartile, 0.89 (95% CI: 0.81–0.98) for the low quartile, 0.88 (95% CI: 0.79–0.96) for the medium quartile, and 0.77 (95% CI: 0.69–0.85) for the high quartile, according to adjusted Model 4 (Ptrend <0.0001). Table 2 displays the multivariate-adjusted hazard ratios and 95% confidence intervals of TPA duration for all-cause mortality.
Table 2.
Associations between total physical activity and all-cause mortality among 109,407 participants.
| TPA groups | No. of cases(%) | No. of men/total participants | Hazard ratio (95% CI) | |||
|---|---|---|---|---|---|---|
| Model 1 | Model 2 | Model 3 | Model 4 | |||
| No TPA | 2456(4.61) | 40,471/53,240 | [Reference] | [Reference] | [Reference] | [Reference] |
|
Quartile 1 (0 < TPA < 75 min) |
592(4.94) | 9743/11,973 | 1.09 (1.01–1.20) | 1.08 (0.99–1.19) | 1.07 (0.98–1.18) | 1.08 (0.98–1.19) |
|
Quartile 2 (75 < = TPA < 360 min) |
530(3.52) | 12,299/15,043 | 0.89 (0.81–0.98) | 0.89 (0.81–0.98) | 0.88 (0.80–0.97) | 0.89 (0.81–0.98) |
|
Quartile 3 (360 < = TPA < 720 min) |
524(3.65) | 11,626/14,366 | 0.87 (0.79–0.96) | 0.86 (0.78–0.95) | 0.86 (0.78–0.95) | 0.88 (0.79–0.96) |
|
Quartile 4 (TPA > 720 min) |
469(3.17) | 12,228/14,785 | 0.81 (0.73–0.89) | 0.77 (0.69–0.85) | 0.77 (0.69–0.85) | 0.77 (0.69–0.85) |
| P value for trend | < 0.0001 | < 0.0001 | < 0.0001 | < 0.0001 | ||
Abbreviation: TPA total physical activity; CI confidence interval.
Model 1 adjusted for age and gender. Model 2 adjusted for Model 1 variables plus cigarette intake; drinking habits; screen time, marriage status, education level, income level. Model 3 adjusted for Model 2 variables plus body mass index (calculated as weight in kilograms divided by height in meters squared); history of Parkinson disease, heart failure, stroke, fatty liver, myocardial infarction. Model 4 adjusted Model 3 variables plus work intensity and work type.
Total physical activity volume per week was associated with a lower risk of mortality when adjusted for sociodemographic factors, lifestyle factors and disease history. According to the analysis using penalized splines, participants who engaged in any amount of TPA had a lower risk of mortality than those who did not engage in any TPA. There was a steady decline in hazard ratios (HRs) with more minutes of TPA per week, up until approximately 2400 min/wk, in relation to all-cause mortality in the fully adjusted model (Fig. 2). The results indicated increased uncertainty beyond 2400 min/wk, as shown by the wide 95% confidence intervals (CIs). At higher levels of TPA, the curve showed non-significant HRs and wide CIs (Fig. 2).
Fig. 2.
Dose-response association between total physical activity per week and all-cause mortality. The y-axis with the shared area representing hazard ratios and 95%CIs. The reference group is physically inactive participants (no PA). Model is adjusted for age, gender, cigarette intake; drinking habits; screen time, marriage status, education level, income level, body mass index (calculated as weight in kilograms divided by height in meters squared), history of Parkinson disease, heart failure, stroke, fatty liver, myocardial infarction plus work intensity and work type.
Subgroup analysis
In the subgroup analysis stratified by these covariates (see eFig. 1), the effect size of TPA on mortality was similar among participants over 65 years of age (HR: 0.91, 95% CI: 0.88–0.94) and participants under 65 years of age (HR: 0.92, 95% CI: 0.90–0.95). Non-significant differences were found for sex (men vs. women), work type (white-collar vs. blue-collar), and history of stroke, fatty liver, and heart failure (eFig. 1). However, work status (employed vs. retired) showed a significant difference in mortality (P for interaction: 0.004). We also examined the combined effect of age and the TPA on all-cause mortality, revealing that as age increased, the association between the TPA and all-cause mortality strengthened (eFig. 2).
Sensitivity analysis
Sensitivity analyses were conducted to confirm the findings. First, after excluding participants who died in the first two years of follow-up, moderate to high levels of TPA were significantly associated with all-cause mortality (HRs: 0.88, 95% CIs: 0.79–0.98 for medium; HRs: 0.84, 95% CIs: 0.75–0.94 for high), with a Ptrend < 0.0001 (see eTable 1 for details). Second, after excluding participants with PD, stroke, heart failure, fatty liver and myocardial infarction, the results remained consistent. TPA at low (HRs: 0.86, 95% CIs: 0.76–0.97), medium (HRs: 0.89, 95% CIs: 0.79–1.01), and high levels (HRs: 0.74, 95% CIs: 0.64–0.85) was significantly associated with all-cause mortality (Ptrend < 0.0001) (see eTable 2 for details). When participants who were followed for less than five years were removed, the mortality rate was lower than that in the main analyses, and the associations between TPA and mortality were still consistent with those in the main analyses (eTable 5). Supportive sensitivity analysis was performed by calculating metabolic equivalent minutes (MET-minutes) to examine the dose response relationship between physical activity and mortality, the results showed that low MET-minutes was positively associated with mortality, with the increase of MET-minutes, mortality decreases significantly (eTable 7). We also examined the collinearity between vigorous physical activity, moderate physical activity, and walking activity. The results showed a strong collinearity between vigorous and moderate physical activity (β = 0.68), while walking exercise had low collinearity with both vigorous (β = 0.31) and moderate (β = 0.33) physical activity (see eTable 3 for details). Furthermore, additional analyses were conducted to investigate the associations between vigorous physical activity, moderate physical activity, walking activity, BMI, and call-cause mortality. No significant associations between BMI, TPA, and mortality were observed (see eTable 4 for details).
Discussion
In this large prospective study, we included a sample of 109,407 Chinese adults with a mean age of 53.3 years. We found that those who engaged in a total physical activity (>75 min/wk) had a reduced risk of all-cause mortality. Our findings suggest that even a small amount of physical activity benefits mortality reduction, which aligns with global guidelines1. Specially for frail individuals, even a lower dose of physical activity may be beneficial18. The dose‒response relationship, observed L-shaped curves in studies conducted among older adults19. We found a steady decline in mortality rates, with more PA minutes accrued, up to approximately 2,400 min/wk, beyond which rates leveled off. After accounting for age, the associations between weekly PA minutes and all-cause mortality increased, suggesting that age is an important covariate in this population.
Many previous studies have examined the associations between PA and all-cause mortality. A cohort study involving 272,550 participants examined the amount of physical activity in metabolic equivalent of task (MET) hours per week. The study found significant associations between engaging in at least 15 MET hours (approximately 300 min/wk) of any activity and the risk of mortality in older individuals20. A harmonized meta-analysis including over 1 million men and women reported that high levels of moderate-intensity physical activity (approximately 60–75 min per day) appear to reduce the increased risk of death21. A systematic review and meta-analysis compiled evidence suggesting that higher levels of physical activity are linked to a lower risk of all-cause mortality. The pooled hazard ratios for all-cause mortality were 0.42 (CI = 0.34, 0.53) for total physical activity, 0.43 (CI = 0.35, 0.53) for moderate-to-vigorous physical activity, and 0.58 (CI = 0.43, 0.80) for light physical activity22.
The present study provides a unique contribution to the dose‒response relationship between total physical activity duration and all-cause mortality, as limited data are available for the Chinese population. When finer gradations were analyzed via spline analysis, the dose‒response curve was J-shaped from a study conducted in Japan11. However, the dose‒response curves varied slightly due to measurement error and the different baseline characteristics of the participants. Discrepancies in curve shapes may reflect methodological differences in PA assessment and/or population characteristics. For example, one study reported an L-shaped curve among participants with major chronic disease23. A single measure of PA at baseline was likely to show an inverted J-shaped or L-shaped dose‒response association24. The J-shaped patterns often emerge in younger/healthy adults25, whereas L-shaped curves typically dominate older populations26. This is because younger population almost have higher level of PA compare to older adults.
In the present study, we determined the maximum safe cutoff points for physical activity. The most commonly used cutoff value for moderate PA is 150 min/wk, and for vigorous PA, it is 75 min/wk27,28. Scott and colleagues published a study in The Lancet with thresholds of 150–750 min/wk (moderate PA) and >750 min/wk (high physical activity), which revealed significant associations with a graded reduction in mortality (HRs: 0.80, 95% CI 0.74–0.87 and 0.65, 0.60–0.71)29. Another study suggested that the most effective hypothetical volume of physical activity was to increase weekly PA to >300 min (risk ratio (RR), 0.66 (0.46–0.86))30. Our findings particularly identified 75 to 360 min/wk, 260 to 720 min/wk and >720 min/wk as significant thresholds for PA and all-cause mortality.
Previous studies have identified various factors related to physical activity and its impact on all-cause mortality, with age being a significant factor. Existing research has emphasized the connection between physical activity and mortality in middle-aged and older adults7. Compared with low levels of PA, moderate to high levels of PA have been linked to increased life expectancy, with a median age of 58 years31. Additionally, engaging in physical activity early in life has been shown to decrease the risk of all-cause mortality32. Diet is a known modifier of health outcomes and may interact with physical activity (e.g., attenuating or amplifying effects)33. While our results are robust across multiple sensitivity analyses, residual confounding from unmeasured variables (e.g., diet, socioeconomic status) may persist.
The present study has several strengths. This was a prospective cohort study in which the association between physical activity and the risk of mortality could be determined in the Chinese population. Potential confounding factors of physical activity and mortality were adjusted for in both the main analysis and sensitivity analysis. Data collection and process control were performed under rigid supervision.
Limitations warrant disclosure. First, the PA data collection was derived from a self-reported survey, of which potential recall bias was considered. Additionally, single-item self-report measures generally underestimate sedentary time compared with device measures34. Second, our study participants were predominantly older male adults, which may limit the generalizability of the findings to other age groups. Third, even though the number of included participants was relatively large, the results may not be fully representative of the entire Chinese population. The characteristics of the participants in our study represent only the low to middle income of Chinese society15. Fourth, information on cause-specific mortality was not available, so caution is advised when interpreting our results, as certain diseases or conditions may be closely linked to physical activity ability. Fifth, our study excluded 14,343 (11.59%) participants due to missing data, which may limit the generalizability of our findings. Furthermore, nearly half of the included population reported no activity, indicating that half of the total population was physically inactive. Previous studies have shown that the association between physical activity and all-cause mortality may vary among physically active and physically inactive individuals35. Although our inactivity prevalence is a bit higher than the national benchmarks, residual misclassification may exist owing to self-reported measurement of physical activity.
Conclusions
Higher levels of total TPA reduced the risk of all-cause mortality by at least 75 min per week, following an inversely nonlinear dose‒response pattern. This study demonstrated that engaging in physical activity can greatly reduce the risk of all-cause mortality, and even low doses of physical activity significantly reduce all-cause mortality risk. These findings support public health strategies that promote incremental increases in movement, particularly for elderly populations. This study supports a paradigm shift in preventive health messaging toward inclusive, achievable physical activity targets while maintaining evidence-based recommendations for optimal duration and intensity.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
We are grateful to thank all participants in this study. Thanks to Dr. Zhongyi Yu and Dr. Lili Huang for consultation.
Author contributions
Study conception and design: FW, XG; data collection: YW, SC, SW; data analysis and interpretation of results: FW, KW, YC; draft manuscript preparation: FW, YL.
Funding
Fudan University startup grant (JIF201036Y).
Data availability
Data available on request from the first author FW.
Declarations
Competing interests
The authors declare no competing interests.
Consent for publication
All authors reviewed the results and approved the final version of the manuscript.
Ethical approval and consent to participate
The study was approved by the Ethics Committees of Kailuan General Hospital. All participants provided written informed consent.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Feifei Wang, Email: feifei.wang@zjxu.edu.cn.
Shuohua Chen, Email: csh01062011@163.com.
Xiang Gao, Email: xiang_gao@fudan.edu.cn.
References
- 1.Bull, F. C. et al. World Health Organization 2020 guidelines on physical activity and sedentary behaviour. [DOI] [PMC free article] [PubMed]
- 2.Belfiore, P., Miele, A., Gallè, F. & Liguori, G. Adapted physical activity and stroke: a systematic review. J. Sports Med. Phys. Fit.58 (12), 1867–1875. 10.23736/S0022-4707.17.07749-0 (2018). [DOI] [PubMed] [Google Scholar]
- 3.McTiernan, A. et al. Physical activity in cancer prevention and survival: A systematic review. Med. Sci. Sports Exerc.51 (6), 1252–1261. 10.1249/MSS.0000000000001937 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Elagizi, A. A Review of Obesity, Physical Activity, and Cardiovascular Disease. (2020). [DOI] [PubMed]
- 5.Wang, F. & Boros, S. The effect of physical activity on sleep quality: a systematic review. (2019).
- 6.Kandola, A. Moving to beat anxiety: epidemiology and therapeutic issues with physical activity for anxiety. Curr Psychiatry Rep. (2018). [DOI] [PMC free article] [PubMed]
- 7.Mok, A., Khaw, K. T., Luben, R., Wareham, N. & Brage, S. Physical activity trajectories and mortality: population based cohort study. BMJ365, l2323. 10.1136/bmj.l2323 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.DiPietro, L. et al. Advancing the global physical activity agenda: recommendations for future research by the 2020 WHO physical activity and sedentary behavior guidelines development group. Int. J. Behav. Nutr. Phys. Act.17 (1), 143. 10.1186/s12966-020-01042-2 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Samitz, G., Egger, M. & Zwahlen, M. Domains of physical activity and all-cause mortality: systematic review and dose-response meta-analysis of cohort studies. Int. J. Epidemiol.40 (5), 1382–1400. 10.1093/ije/dyr112 (2011). [DOI] [PubMed] [Google Scholar]
- 10.Wen, C. P. et al. Minimum amount of physical activity for reduced mortality and extended life expectancy: a prospective cohort study. Lancet Lond. Engl.378 (9798), 1244–1253. 10.1016/S0140-6736(11)60749-6 (2011). [DOI] [PubMed] [Google Scholar]
- 11.Seino, S. et al. Dose-Response Associations of Physical Activity and Sitting Time with All-Cause Mortality in Older Japanese Adults. J Epidemiol Published online Dec.24, 10.2188/jea.JE20220246 (2022). [DOI] [PMC free article] [PubMed]
- 12.Lu, S., Shuai, Z. & Lu, Y. The dose-response relationship between physical activity and the risk of death from pneumonia in middle-aged and older adults: A meta-analysis. Med. (Baltim).103 (21), e38220. 10.1097/MD.0000000000038220 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Sun, J., Wu, H., Zhao, M., Magnussen, C. G. & Xi, B. Dose-response association of leisure time physical activity with mortality in adults with major chronic diseases. Front. Nutr.9, 1048238. 10.3389/fnut.2022.1048238 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Yu, Y. et al. The cardiovascular and cerebrovascular health in North China from 2006 to 2011: results from the KaiLuan study. Front. Cardiovasc. Med.8, 683416. 10.3389/fcvm.2021.683416 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Liu, X. et al. Cumulative exposure to ideal cardiovascular health and incident diabetes in a Chinese population: the Kailuan study. J. Am. Heart Assoc.5 (9), e004132. 10.1161/JAHA.116.004132 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Qu, N. & Li, K. ji. [Study on the reliability and validity of international physical activity questionnaire (Chinese Vision, IPAQ)]. Zhonghua Liu Xing Bing Xue Za Zhi Zhonghua Liuxingbingxue Zazhi. 25(3):265–268. (2004). [PubMed]
- 17.Yang, W. et al. Different levels of physical activity and risk of developing type 2 diabetes among adults with prediabetes: a population-based cohort study. Nutr. J.23 (1), 107. 10.1186/s12937-024-01013-4 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Martinez-Gomez, D. et al. Physical activity and All-Cause mortality by age in 4 multinational megacohorts. JAMA Netw. Open.7 (11), e2446802. 10.1001/jamanetworkopen.2024.46802 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Fukushima, N. et al. Dose-Response Relationship of Physical Activity with All-Cause Mortality among Older Adults: An Umbrella Review. J Am Med Dir Assoc. S1525-8610(23)00835-6 (2023). 10.1016/j.jamda.2023.09.028 [DOI] [PubMed]
- 20.Watts, E. L. et al. Association of leisure time physical activity types and risks of All-Cause, Cardiovascular, and cancer mortality among older adults. JAMA Netw. Open.5 (8), e2228510. 10.1001/jamanetworkopen.2022.28510 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ekelund, U. et al. Does physical activity attenuate, or even eliminate, the detrimental association of sitting time with mortality? A harmonised meta-analysis of data from more than 1 million men and women. Lancet Lond. Engl.388 (10051), 1302–1310. 10.1016/S0140-6736(16)30370-1 (2016). [DOI] [PubMed] [Google Scholar]
- 22.Liew, S. J., Petrunoff, N. A., Neelakantan, N., van Dam, R. M. & Müller-Riemenschneider, F. Device-Measured physical activity and sedentary behavior in relation to cardiovascular diseases and All-Cause mortality: systematic review and Meta-Analysis of prospective cohort studies. AJPM Focus. 2 (1), 100054. 10.1016/j.focus.2022.100054 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Arem, H. et al. Leisure time physical activity and mortality: a detailed pooled analysis of the dose-response relationship. JAMA Intern. Med.175 (6), 959–967. 10.1001/jamainternmed.2015.0533 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Lee, D. H. et al. Physical activity and all-cause and cause-specific mortality: assessing the impact of reverse causation and measurement error in two large prospective cohorts. Eur. J. Epidemiol.36 (3), 275–285. 10.1007/s10654-020-00707-3 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Bhaskaran, K., Dos-Santos-Silva, I., Leon, D. A., Douglas, I. J. & Smeeth, L. Association of BMI with overall and cause-specific mortality: a population-based cohort study of 3·6 million adults in the UK. Lancet Diabetes Endocrinol.6 (12), 944–953. 10.1016/S2213-8587(18)30288-2 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.López-Bueno, R., Yang, L., Stamatakis, E. & Del Pozo Cruz, B. Moderate and vigorous leisure time physical activity in older adults and alzheimer’s disease-related mortality in the USA: a dose-response, population-based study. Lancet Healthy Longev.4 (12), e703–e710. 10.1016/S2666-7568(23)00212-X (2023). [DOI] [PubMed] [Google Scholar]
- 27.Zhao, M., Veeranki, S. P., Magnussen, C. G. & Xi, B. Recommended physical activity and all cause and cause specific mortality in US adults: prospective cohort study. BMJ370, m2031. 10.1136/bmj.m2031 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Stamatakis, E. et al. Association of wearable device-measured vigorous intermittent lifestyle physical activity with mortality. Nat. Med.28 (12), 2521–2529. 10.1038/s41591-022-02100-x (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Lear, S. A. et al. The effect of physical activity on mortality and cardiovascular disease in 130 000 people from 17 high-income, middle-income, and low-income countries: the PURE study. Lancet Lond. Engl.390 (10113), 2643–2654. 10.1016/S0140-6736(17)31634-3 (2017). [DOI] [PubMed] [Google Scholar]
- 30.Yang, Y. I. et al. Mortality effects of hypothetical interventions on physical activity and TV viewing. Med. Sci. Sports Exerc.53 (2), 316–323. 10.1249/MSS.0000000000002479 (2021). [DOI] [PubMed] [Google Scholar]
- 31.Chudasama, Y. V. et al. Physical activity, multimorbidity, and life expectancy: a UK biobank longitudinal study. BMC Med.17 (1), 108. 10.1186/s12916-019-1339-0 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Feter, N. et al. Physical activity during early life and the risk of all-cause mortality in midlife: findings from a birth cohort study. Eur. J. Public. Health. 33 (5), 872–877. 10.1093/eurpub/ckad084 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Bull, E. R. et al. Interventions to promote healthy Eating, physical activity and smoking in Low-Income groups: a systematic review with Meta-Analysis of behavior change techniques and Delivery/Context. Int. J. Behav. Med.25 (6), 605–616. 10.1007/s12529-018-9734-z (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Prince, S. A. et al. A comparison of self-reported and device measured sedentary behaviour in adults: a systematic review and meta-analysis. Int. J. Behav. Nutr. Phys. Act.17 (1), 31. 10.1186/s12966-020-00938-3 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Stamatakis, E. et al. Sitting Time, physical Activity, and risk of mortality in adults. J. Am. Coll. Cardiol.73 (16), 2062–2072. 10.1016/j.jacc.2019.02.031 (2019). [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data available on request from the first author FW.