In China, the number of patients with myocardial infarction (MI) has been dramatically increasing, and is estimated to reach 23 million by 2030 [1]. MI compromises cardiac function and remodels the cardiac structure, and thus cardiac structure and function may vary in different post-MI periods [2, 3, 4]. The mortality risk of MI survivors is still high [5], and varies at different post-MI periods (within and beyond 1 year) [6]. Physical activity (PA) has been shown to reduce all-cause mortality and cardiovascular mortality of MI patients by 8%–37% and 7%–38%, respectively [7]. Aerobic PA and other lifestyle interventions have been reported to reduce coronary atherosclerosis to different degrees within 1 and 5 years [8]. And PA and exercise-based cardiac rehabilitation have been recommended as important components of secondary prevention of MI and cardiovascular disease (CVD) [9, 10, 11, 12, 13]. The guideline for cardiac rehabilitation in China recommends regular aerobic PA within 1 year after MI and individualized training beyond 1 year after discharge [14]. The possible mechanisms of mortality reduction associated with PA may include improved cardiorespiratory fitness and cardiac remodeling [15]. Thus, we hypothesized that PA may have a different impact on mortality reduction among MI survivors at different post-MI periods.

However, evidence is lacking on the association between different volumes and patterns of total PA and the mortality risks at different post-MI periods. First, there is limited evidence on the association between PA collected at different post-MI periods and mortality among MI survivors. Prior studies primarily used longitudinal cohort follow-up through repeated surveys [16, 17, 18, 19, 20], and focused on PA at 6–10 weeks post-MI [19], or the change of levels of PA after MI [16, 17, 18, 20]. Second, there is a lack of evidence on non-leisure time PA (household, occupational, and transport PA) or total PA (both leisure and non-leisure time PA) among MI patients. Specifically, PA was mainly performed during non-leisure time rather than leisure time in China and other low- and middle-income countries [21]. This differed greatly from those in high-income countries, and the previous evidence mainly focused on the high-income ones [22]. Besides, previous analyses mainly focused on certain populations with small sizes [17, 18, 20]. A comprehensive understanding of the time-varying effect of physical activity on mortality is needed in clinical practice for developing a more tailored strategy for secondary prevention after MI.

We analyzed data from the China Patient-centered Evaluative Assessment of Cardiac Events (PEACE) Million Persons Project (MPP), a population-based cohort study covering 31 provinces in mainland China, to explore the time-varying effect of different volumes and patterns of total PA (including leisure and non-leisure time PA) on mortality risk among MI survivors.

Our study found that sufficient total PA was associated with lower risk of mortality among MI survivors. The dose-response association of total PA and mortality was curvilinear within 1 year and linear after 1 year of MI onset. The reduction in mortality risk among MI survivors appeared to be apparent with moderate PA within 1 year after MI, and those with high PA beyond 1 year after MI. Both leisure and non-leisure patterns were associated with lower mortality risk, independent of the post-MI periods.

This study adds to the prior literature in several ways. First, we provided information regarding the association between total PA and mortality risk among MI survivors and found a difference between the within-1-year and beyond-1-year groups. Guidelines for cardiac rehabilitation recommend regular moderate-intensity aerobic PA lasting 30–60 minutes on each day and 3–5 days/week for stage 2 rehabilitation (within 1 year after discharge), and an individual home-based training program for stage 3 rehabilitation (beyond 1 year after discharge) [14]. One possible reason for the different recommendations is the potential heterogeneity in cardiac structure and function [3, 4], and mortality risk [5] between two post-MI periods. In this study, a significant difference in the distribution of PA was observed among MI survivors within and beyond 1 year of MI onset. Within 1 year, MI survivors tended to perform less leisure time PA, compared with those beyond 1 year. It is a reasonable assumption that MI survivors within 1 year tended to be reluctant to perform leisure time PA for fear of adverse events following their acute MI. However, although they did not actively select leisure time PA, they performed more non-leisure time PA and total PA, and had a lower proportion of insufficient PA. Apart from the possible different cardiac structures and functions, we speculate that the difference in the association between PA and mortality risk at two post-MI periods might be partly due to the distribution of PA habits.

Second, we report the association between different volumes of PA and the risk of mortality at different post-MI periods. Previous studies of graded effects of total PA were limited to the general population and patients with CVD [22, 33]. Although the Global Burden of Disease study recommends 3000–4500 MET minutes/week of total PA [28] based on a meta-analysis quantifying the dose-response associations in five chronic diseases (including ischemic heart disease events) [34], it is still unclear whether a similar volume could also work in MI survivors at different post-MI periods. In our study, $>$3000 MET minutes/week of total PA was associated with lower mortality risk in the within-1-year group, while in the beyond-1-year period, a higher volume ($>$4500 MET minutes/week) of total PA was associated with lower risk of mortality. This curvilinear association was found in patients within 1 year of MI onset, with a steep risk reduction at a moderate volume of total PA and risk reduction plateauing at higher volumes, which suggested no additional health benefits above a certain volume of total PA. The linear relationship in the beyond-1-year group indicates that performing PA at a higher volume was consistently associated with lower risk of mortality. The potential explanation for the different shapes of dose-response association in different post-MI periods is not clear. Previous studies reported a similar non-linear shape of the dose-response association between PA and mortality in individuals without CVD [33] and in those with stable coronary heart disease [35], while a the linear relationship was reported in CVD patients [33]. Previous evidence showed that vigorous PA transiently increased the risk for sudden cardiac death [36], and vigorous intensities and large PA volumes were both associated with potential accelerated coronary artery calcification, exercise-induced cardiac biomarker release, myocardial fibrosis, and atrial fibrillation [37]. We speculate that cardiac structure and function are not sufficiently recovered within 1 year of MI onset, and that the difference in cardiac structure and function may relate to the different dose-response associations.

Third, our study filled the gap in the association between different patterns of sufficient PA and mortality risk. We found that sufficient total PA, either leisure (higher proportion of leisure time PA) or non-leisure (higher proportion of non-leisure time PA) pattern, was associated with lower risk of mortality, compared with insufficient total PA. In our study, only a few MI patients had leisure pattern. This is consistent with previous studies which observed that PA in China mainly involved occupation and housework, with less contribution from leisure time PA [21]. Prior evidence on leisure time PA [33, 38] and non-leisure time PA [22, 33] was reported in the general population and patients with CVD, among which the evidence was inconsistent on non-leisure time PA. And in the research where non-leisure time PA was categorized into the domains of household, occupational, and transport, the findings were still inconsistent [39, 40, 41, 42]. Leisure time PA is usually a short period PA with a long recovery time, which may improve cardiovascular risk factors (such as insulin resistance, hypertension, dyslipidemia, and obesity) [40, 43]. However, non-leisure time PA, such as occupational PA, is of low intensity or long duration and may not improve cardiovascular health. Occupational PA could raise 24-hour heart rate and blood pressure, and increase the levels of inflammation [41]. While the potential mechanisms of household and transport PA still need further study. The possible influential factors are that some housework is also an aerobic PA with sufficient recovery time, while transport PA may be affected by traffic and harmful exposure (e.g., gases on roads) as well as active or passive commuting types [42].

Our findings have several important implications. First, appropriately high PA after MI indicates lower risk of mortality. This finding serves as empirical support for making scientifically warranted secondary prevention recommendations for the post-MI population to promote sufficient PA. Second, the evidence on proper volumes and patterns of PA for patients at different post-MI periods could facilitate more tailored health advice for PA after MI. Third, for MI survivors within 1 year of their MI onset, as non-leisure time PA is usually performed on a daily basis, we suggest selecting an adequate volume of leisure time PA after fully considering their non-leisure time PA to reach a total level of approximately at least 3000 MET minutes/week within 1 year of MI onset, and no less than 4500 MET minutes/week beyond 1 year of MI onset. Nevertheless, patients should be cautious about their choice in the proper volume of PA. The volumes well above the PA recommendations were reported to be associated with increased mortality risk in CVD patients [44, 45]. More importantly, MI patients should consult with their physicians before adjusting for their PA volume.

Our study has several limitations. First, MI history and PA information were self-reported, which may bring recall bias and may overestimate PA volume [46]. However, questionnaires were collected by trained professionals following a standardized process, the self-reported medical information was based on physician diagnosis, and the questions of PA information had been used and validated in other cohort studies [24, 25, 26]. We assumed that the recall bias is relatively small in a large-scale population. Second, we used the recommended total PA level of the Global Burden of Disease study rather than the widely used World Health Organization guideline [47]. In Chinese adults, PA mainly involves occupation and housework, without much input from leisure time PA [21], so the studies in China usually referred to the World Health Organization standard for leisure time PA rather than total PA [48]. Third, the study lacks data on important clinical risk factors, such as exercise capacity, cardiac function, and disease severity, which are also potential confounders. Nevertheless, we adjusted for history of heart failure, history of chronic kidney disease, metabolic factors, lifestyle factors, and medication use to reduce the confounding effect when analyzing the association between PA and mortality. Fourth, a risk of reverse causality is present, so participants with less severity of MI may have a higher PA and a better prognosis. To address the concerns related to reverse causality, we excluded the participants who reported problems in walking, self-care, and daily activities in the EQ-5D questionnaire. In addition, we did sensitivity analyses using the inverse probability weighting method, and the results did not substantially change our main analyses. However, reverse causality may be unavoidable in observational studies, establishing the optimal volume and pattern of PA in each clinical setting would require randomized head-to-head comparisons.

Our findings demonstrated that sufficient total PA was associated with lower risk of mortality compared with insufficient PA among MI survivors. The dose-response association varied by the different post-MI periods. Both leisure and non-leisure patterns were associated with lower mortality risk, independent of the post-MI periods. These findings could facilitate tailored recommendations on PA in secondary prevention after MI. Further studies are needed to confirm the causal effect of different volumes and patterns of PA on mortality risk at different periods after MI.

CVD, cardiovascular disease; China PEACE MPP, China Patient-centered Evaluative Assessment of Cardiac Events Million Persons Project; CI, confidence interval; HR, hazard ratio; ICD-10, International Classification of Diseases, 10th edition; IQR, interquartile range; MI, myocardial infarction; MET, metabolic equivalent of task; PA, physical activity; SD, standard deviation.

The data that support the findings of this study are available from China Patient-centered Evaluative Assessment of Cardiac Events Million Persons Project (China PEACE MPP) but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of China PEACE MPP. China PEACE MPP only provides conditional data access for qualified researchers with legitimate requests; a formal application and research proposal is required. Please contact cvd-project@nccd.org.cn to seek approval for data access.

XZ, XL and TL designed the research study. XL conceived of the China Patient-centered Evaluative Assessment of Cardiac Events Million Persons Project (China PEACE MPP) and takes responsibility for all aspects of it. XiaYZ, YY, JLC, WX, LJS, HY, WYH and YZ participated in the project operation. TL and XinYZ analyzed the data. TL wrote the manuscript, with further contributions from XZ, XL, XinYZ, XLW, JLS, CQW, AXT, XiaYZ, YY, JLC, WX, LJS, HY, WYH, and YZ. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript.

The project protocol was approved by the central ethics committee at Fuwai Hospital, Beijing, China (Approval No. 2014-574). Written informed consent was obtained from all enrolled participants.

We appreciate the multiple contributions made by study teams at the National Center for Cardiovascular Diseases, and the local sites in the collaborative network in the realms of study design and operations.

This research was funded by the CAMS Innovation Fund for Medical Sciences (CIFMS), grant number 2021-I2M-1-011; and National High Level Hospital Clinical Research Funding, grant number 2022-GSP-GG-4.

The authors declare no conflict of interest.