Sense of coherence predicts post-myocardial infarction trajectory of leisure time physical activity: a prospective cohort study

Physical activity confers a survival advantage on post-myocardial infarction (MI) patients [1–4]. Regularly active MI survivors were shown to have approximately half the risk of dying compared with inactive patients, irrespective of pre-MI habits [5]. Despite this clear advantage, a minority of post-MI patients are regularly active [5, 6]. It remains unclear what differentiates active from non-active post-MI patients.

Sense of coherence (SOC), a central construct of the salutogenic model of health, has been associated with health outcomes. This model was designed to explain improvement in one's location on the health-disease continuum [7], with SOC representing a person's confidence that they have the resources to cope with problems and challenges [8]. Previous research has reported a link between a strong SOC and reduced mortality from cancer and CVD [9, 10], suggesting that SOC confers a form of resilience against disease. The explanatory pathway for this relationship remains unclear, although research has indicated that health behaviours may explain some of the association [10].

In a previous study, we found that post-MI patients with a stronger SOC were more likely to quit smoking than those with a weak SOC [11]. In the current study, we examined whether a similar relationship exists with post-MI physical activity. By elucidating which factors influence the adoption of important secondary prevention behaviours including physical activity, post-MI care can be improved. We therefore examined the predictive role of SOC in the long-term trajectory of LTPA in a community cohort of post-MI patients followed up for 13 years.

SOC clearly predicted LTPA throughout follow-up. Point prevalence of regular activity in the lower versus upper tertiles was 37-42% versus 50-60% during T2-T5 (Figure 1). Inactivity was reported in 33-39% of the lower SOC tertile, versus 20-22% of the upper tertile (Figure 2).

Patients with a weaker SOC exhibited less dramatic changes in LTPA levels after MI compared with those with a stronger SOC. While among strong SOC patients point prevalence of regular activity improved by approximately 20% during follow up, it improved by around 10% in the lower SOC patients (Figure 1).

In this cohort of post-MI patients, the predictive role of SOC in long-term trajectory of LTPA was evident, even after controlling for sociodemographic and psychological factors and disease severity. Weak SOC was consistently associated with lower engagement in LTPA throughout follow-up. SOC was associated with income, educational level, employment status, self-rated health and depressive symptoms, matching the findings from a healthy population survey [20], and indicating that SOC reflects personal, social and economic resources.

Previous research has investigated the role of SOC in health outcomes and behaviours. The European Prospective Investigation into Cancer (EPIC) study reported a 30% reduction in both all-cause and cardiovascular mortality associated with a strong SOC, independent of age, sex, and prevalent chronic disease [9]. This link suggests that SOC confers a form of resilience against disease, however the explanatory pathway for the SOC-mortality relationship remains unclear. In a follow-up of the EPIC, lifestyle choices including diet, smoking and physical activity, and SES explained 23% of the association between SOC and mortality [10]. Since lifestyle behaviours were measured at a single time point, their true contribution may be even greater. The current study measured physical activity on 5 separate occasions over a 13 year period and found a significant association between SOC at baseline and LTPA trajectory.

Evidence for the link between SOC and health behaviours has been reported in several studies, including in healthy college students [21] and hypertensive patients [22]; SOC has previously been associated with smoking [22, 23] and oral health behaviours [24–26], as well as physical activity [10, 23, 24, 27]. In a British survey of 18,000 adults, participants with the strongest SOC were 28% less likely to be current smokers and 36% less likely to be physically inactive than those with the weakest, independent of sociodemographics [23], although this study was based on a 3-item assessment of SOC rather than the full 29-item questionnaire used in the present study. In a student cohort followed up for 3 years, frequency of physical activity was related to strength of SOC [27]. A prospective study of middle-aged Finnish men found that differences in LTPA depended on SOC [28]. However, these studies were cross-sectional [23, 24], involved small samples [27], or assessed SOC several years after physical activity [28] therefore providing limited evidence of the predictive role of SOC in physical activity.

The relationship between SOC and health behaviours has rarely been examined in MI patients. A small-scale study tackled this issue, and found that post-MI patients who chose to participate in an exercise-based rehabilitation programme had significantly higher SOC scores at baseline than those who declined to take part [29]. Additionally, a discriminant analysis showed that SOC score successfully differentiated attendees from non-attendees. The current study examined the predictive role of SOC in engagement in physical activity in a much larger sample, controlled for a wide range of confounding variables, and had a follow-up of 13 years. We focused on leisure time physical activity since, although cardiac rehabilitation programmes are indispensable in post-MI recovery, long-term maintenance of physical activity is key to secondary prevention.

Our finding that baseline SOC score was significantly associated with LTPA after 10-13 years suggests that a strong SOC may be involved in the maintenance of health behaviours. Figure 1 illustrates that LTPA prevalence increased in all groups after MI, although the increase was steeper in strong SOC patients. The problem encountered, as with all health behaviours, is the gulf between adoption and maintenance, indeed Leung and colleagues found in a cohort of patients with coronary artery disease that while exercise behaviour increased after discharge from hospital, it had decreased after 18 months [30]. It is therefore interesting to note that while many of our post-MI cohort increased their activity level after MI, no doubt in many cases on medical advice, it was the individuals with a weak SOC who had much higher odds of decreasing levels of activity 10-13 years later. It should also be noted that at the last interview (T5), less than 1% of LTPA data were missing, therefore a more accurate picture of activity habits was available compared to earlier timepoints which had 10-15% missing data.

A strong SOC was additionally associated with pre-MI LTPA habits, perhaps indicating that these individuals led a healthier lifestyle to begin with. We may then have expected that high SOC patients would remain consistently ahead of their lower SOC counterparts in the activity stakes. The fact that LTPA not only remained more prevalent, but increased at a higher rate post-MI in the higher compared to lower SOC groups, suggests that SOC may be related both to health behaviours in general, and to secondary prevention behaviours as an adaptive coping mechanism after illness.

SOC was found to predict LTPA trajectory in a cohort of MI survivors. The primary use of this finding is in identifying high-risk MI survivors, with SOC representing a summary measure of personal, social, psychological and economic resources. SOC is easy to ascertain during hospitalization and can help to identify patients who are less likely to incorporate physical activity in their daily life, who may require additional help in adopting preventive health behaviours, which can dramatically improve post-MI prognosis. These findings have implications not only for secondary prevention of cardiovascular disease but for other illnesses and may be applicable to health promotion in general.

This work was supported in part by the Israel National Institute for Health Policy and Health Services Research (grant number r/89/2008 to Drs. Drory and Gerber), and by the Environment and Health Fund, Jerusalem, Israel (grant award number RGA0904 to Dr. Gerber), which had no role in study design, data acquisition, writing of the manuscript or decision to submit for publication.