Physical activity at leisure time and long-term survival among heart failure patients with a history of myocardial infarction: a prospective cohort study

Abstract

Listen

Translate article

Purpose: Few studies have investigated the relationship between leisure time physical activity (LTPA) and outcomes in HF. The contribution of psychosocial factors to this relationship is vital as they can be targeted in preventive interventions. Since depressed, and/or socially unsupported people may have a higher mortality risk, they may gain more from LTPA, but be less active. Aims included studying in patients with a history of myocardial infarction (MI): (1) the relationship between LTPA and mortality, overall and by HF status, and (2) the confounding/moderating effect of depression and social support (SS). Methods: Participants were drawn from the Israel study of First Acute Myocardial Infarction, a multi-center prospective cohort study of patients admitted to hospital with incident MI. Information on LTPA, depression (Mental Health Inventory), SS (Multidimensional Scale of Perceived Social Support) and other prognostic indicators were collected in 2002-2005, 10-13 years after index MI. Follow-up mortality data were obtained through 2011. LTPA (self-reported walking and/or other aerobic activities) was classified as regular (1-150, 150-300, or >300 weekly minutes), irregular, or none. HF was defined according to clinical history and a New York Heart Association classification of III-IV. Cox regressions were performed. Results: The analysis included 1169 patients (mean age 63.9; 82% men); 237 (20%) met HF criteria. During a median of 8.4 years follow-up, 303 deaths occurred (126 (52%) in HF and 181 (19%) in HF-free cases). Regular LTPA was more common in HF-free (47%) vs. HF cases (36%; χ2<0.01). LTPA was positively associated with SS, educational level, participation in cardiac rehabilitation and education on lifestyle factors, and inversely associated with depression and having comorbidity. LTPA showed a strong relationship with mortality after adjusting for demographic, clinical, interventional and psychosocial (including depression and SS) covariates (p for trend<0.01 in all models). The relation did not differ between HF and HF-free cases (p for interaction=0.41); hazard ratios (95% CIs) for death in the most vs. least active categories were 0.43 (0.20-0.90) among HF and 0.47 (0.29-0.76) among HF-free cases. There was no evidence of interaction between depression or SS and LTPA regarding mortality risk. Conclusions: LTPA is a strong predictor of mortality in post-MI patients, regardless of HF status. HF patients, and, among HF patients, those with a high level of depression or a low level of SS are less likely to be active, which defines preventive intervention opportunities.