• Epidemiology
• Risk Factors
• Electronic Health Records
• Health Services

Cardiovascular diseases (CVD) are highly prevalent conditions that carry one of the highest morbidity and mortality burdens worldwide.1 Physical activity is a known protective factor for prevention of non-communicable diseases such as CVD, and for all-cause and cardiovascular mortality. Implementing prevention strategies to promote a physically active lifestyle should therefore be a public health priority.

A recent systematic review suggests that patients with coronary heart disease who maintain or engage in physical activity can significantly reduce mortality risks compared to those inactive over time.2 Changes in physical activity describe trajectories that have the potential of better defining individual risk profiles compared to single-time assessments.2 Recently, growing interest has been drawn to the possibility of achieving cardiovascular benefits by engaging in physical activity also in late-life.3 4 For most non-communicable disease risk factors, longer duration of exposure is associated with higher chances of developing the disease. Similarly, if the exposure to a protective factor is too short, this may not exert any relevant beneficial effect. This could be the case of beginning physical activity in late-life. Numerous studies have shown that increasing physical exercise in mid-life is associated with a significantly lower mortality risk, also among patients with CVD,5 6 but evidence in late-life remainsscarce. Given the presence of ageing populations with a high prevalence of CVD, examining this association has great clinical and public health significance.

A new study by Kang and colleagues7 published in this journal suggests that increasing physical exercise habits after an incident cardiovascular event in late-life can significantly reduce mortality risks. The authors selected data from 558 147 individuals, a 10% sample of the 5.5 million people aged ≥60 years present in the Korean National Health Insurance database, and included only those with clinical assessments between 2003 and 2012. Incident CVD was defined as any first occurrence of acute myocardial infarction, heart failure, or ischaemic stroke. Only people with a physical activity assessment within 2 years before and after CVD onset were included, and were categorised as persistent non-exercisers, exercise dropouts, new exercisers, and exercise maintainers. The final cohort comprised 6076 individuals, 50.6% men, with a median age of 72 years (IQR 69–76) and 3.8 years of follow-up (IQR 2.2–5.7). All-cause mortality rates and hazard ratios (HR), compared to persistent non-exercisers (incidence rate (IR) 4.8 per 100 person-years), were significantly lower for new exercisers (IR 3.5 per 100 person-years; HR 0.73, 95% CI 0.58 to 0.91) and exercise maintainers (IR 2.9 per 100 person-years; HR 0.53, 95% CI 0.38 to 0.73). Exercise dropouts had a similar all-cause mortality rate (IR 4.6 per 100 person-years) and a non-significant risk difference compared to persistent non-exercisers (HR 0.92, 95% CI 0.70 to 1.22). The same association was present for cardiovascular and non-cardiovascular deaths. When stratifying by age, this association was also found among participants younger and older than 75 years, and regardless of sex or baseline Charlson Comorbidity Index.

Figure 1 reports some of the most relevant cardiovascular and mortality risk factors to outline part of the complexity of lifetime cumulative exposures. This study has focused on late-life exposures, examining physical activity trajectories before and after an incident cardiovascular event, but physical activity patterns, cumulative risk and protective factors could be affected by previous unmeasured early- and mid-life exposures. Nevertheless, these findings suggestthat regardless of what occurred earlier in life, among elderly cardiovascular patients, becoming or maintaining physically active can reduce all-cause mortality by, respectively, 27% and 47%, compared to those inactive over time. This finding is of great public health importance as it suggests physical activity could contribute to significantly reduce the mortality burden among older cardiovascular patients.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Exposures to internal and external protective and risk factors for cardiovascular diseases (CVD), overall, CVD, and non-CVD mortality, along with a schematic representation of physical activity changes in a lifetime approach, and before-and-after incident CVD.

Several mechanisms could underlie the benefits of physical activity in late-life and contribute to explain the observed association; these include improved endothelial function and antiatherosclerotic effects.8 Endothelial dysfunction results in impaired nitric oxide production, abnormal vasoconstriction, inflammation and oxidative stress. Physical activity can promote the production of nitric oxide, and consequent vasodilation. Exercise is also associated with reduced atherogenic lipid profiles, lower blood pressure, and decreased risk of insulin resistance. These are all well-known protective factors for CVD, although their benefits are not only limited to reducing the incidence of the disease, but also improve prognosis. However, an indepth understanding of the underlying pathophysiological mechanisms is still missing.

There are, nonetheless, some limitations inherent to this study. First, the absence of an assessment of the severity of patients’ cardiovascular condition or data on disease progression may represent one of the main confounding factors, which is common among similar observational studies. Exercise changes could be the result of physical impairment caused by the cardiovascular event, and becoming inactive may be a consequence of the clinical course of the disease. Second, the presence of comorbidities among patients with CVD, and especially peripheral artery disease, has been associated with a significantly reduced propensity to perform physical activity as well as with increased mortality. However, fully adjusted results controlled for confounding of peripheral artery disease, as well as numerous other conditions, but the observed association was maintained. Third, CVDs for participant inclusion were a heterogeneous group of conditions (i.e., heart failure, ischaemic stroke, and acute myocardial infarction), but the association between physical exercise changes and all-cause mortality did not vary significantly across these groups (p value for interaction 0.088). Among participants with acute myocardial infarction, the association might have been confounded or mediated by the subsequent onset of heart failure that was not controlled for, but this group was the least numerous and it is unlikely to have affected the overall results. Fourth, the inclusion of individuals with a physical activity assessment after incident CVD might introduce selection bias as healthier participants with a less severe disease were more likely to undergo follow-up examination. Finally, a longer observation period to include participants’ end of life could be warranted.

Despite these limitations, these findings are consistent with previous evidence and their robustness was confirmed by numerous sensitivity analyses. These results are very encouraging and convey an important public health message, suggesting that becoming or maintaining physically active in late-life, even after a cardiovascular event, is not too late to reduce mortality risks. Potential benefits of physical activity might go beyond those reported in the study. In fact, only moderate to vigorous physical activity was considered for exposure assessment, although previous studies have shown that light physical activity is also associated with lower mortality.9

These results support current guidelines recommending that people with chronic conditions engage in regular lifestyle physical activity, according to their abilities.10 Guidelines have been established based on studies largely focusing on younger individuals, so confirming that similar health benefits can be achieved in late-life is especially useful.

Future research based on objectively measured physical activity (i.e., by means of wearable devices) might provide greater insight on the benefits of different intensity levels of activity. Furthermore, monitoring lifetime trajectories, including those in early- and mid-life, could help examine the role of age as a possible effect modifier. Trajectories in late-life alone might not be sufficient to accurately determine an individual’s cumulative physical activity exposure. Despite carrying more information than single-time assessments, trajectories defined by two or three assessments still represent a relatively small portion of lifetime exposure. Online supplemental figure 1 represents all possible combinations of trajectories of physical activity, assuming only two measurements in early-, mid- and late-life, with a dichotomous categorisation (i.e., active or inactive). Even though trajectories would most likely be distributed non-randomly, and a frequency-based grouping could be performed, there is, a priori, potential for a very wide range of exposures. Nevertheless, by covering a cumulative exposure that spans over a longer period of time, trajectories are likely to be more clinically informative than single-time assessments. In the near future, physical activity trajectories could be integrated in clinical practice given their potential prognostic implications and may help to guide clinical decisions.

Further research should aim at confirming the study results in other settings. Non-interventional studies that can control for CVD severity and disease progression to reduce confounding bias, as well as randomised controlled trials that promote an active lifestyle and physical exercise pattern changes, and pathophysiological studies to determine causality of the findings, could contribute to support public health recommendations and healthcare professionals in clinical practice to promote physical activity among older cardiovascular patients.