Discussion
In the REGARDS study population, we found that history of MI was associated with worse function as measured by IADL scores, IADL dependency, ADL scores, chair stands and PCS. These metrics differed in the type of assessment (self-reported or observed) and in the types of activities targeted. However, the results of the current study were largely consistent across metrics, suggesting that MI could lead to impairment in various aspects of physical function. The exceptions were ADL dependency, which was uncommon in this study of community dwelling individuals (<1%), and gait speed, measured using an 8-foot timed walk, which was practical for in-home assessment but is a shorter distance than is recommended for use in clinical practice.25 There was some evidence for stronger associations among women, and particularly black women. In this study, we examined individuals with both incident and recurrent MI. The population with a history of MI includes people with recurrent events and with distant history of MI, but these individuals have not previously been well-represented in the literature on MI and physical function. Although associations tended to be numerically greater for participants with MI during follow-up compared with MI at baseline only, most of these differences did not reach statistical significance. These findings in a contemporary population complement prior studies examining the association of MI with function conducted with ST-segment elevation MI was more common and extend prior work suggesting that black individuals and women, particularly black women, are especially vulnerable to worse physical function associated with MI.
Methods for assessing physical function are variable in clinical practice and in research. Several studies of MI and physical function have examined the ability to conduct ADL and IADL.7 8 Other studies have assessed observed physical performance measures like gait speed and chair stands; self-reported measures of physical health-related quality of life and function like PCS; and other measures of physical function and disability.9–12 A unique aspect of the current study among REGARDS participants was the multicomponent assessment of physical function covering many of the domains examined in prior studies, with small-to-moderate differences observed between participants with and without MI. For the IADL and ADL scores used in this study, a 1-unit difference is interpretable as one additional limitation (having difficulty completing the task) or dependency (requiring another assistance from another person). Although the mean differences in IADL and ADL scores between individuals with and without MI were relatively modest (<1), the odds of IADL dependency was 56% higher among individuals with incident MI compared with individuals without MI and the odds of ADL dependency was 84% higher, though not statistically significant. The clinically meaningful change in time to complete five chair stands has been reported to range from 0.5 to 4.7 s, depending on estimation method.26 In this study population, the adjusted difference between participants with incident MI and those without MI was 1.2 s. The clinically meaningful difference in PCS has been reported to range from 0.6 to 12.127; in this study population, individuals with incident MI had PCS scores that were 4.2 points lower than participants without MI.
Mechanisms that may link MI with physical function include pathologic and psychosocial changes caused by MI, limitations in function increasing risk of MI, effects of hospitalisation and acute and postacute therapies for MI and shared risk factors between MI and worse function.28 While this study and most prior studies have found that MI and other coronary heart diseases are associated with worse function, the Medical Outcomes Study found that there was substantial variation in function within groups of individuals with chronic conditions, including heart disease.12 In the biopsychosocial model of function and disability, physical function results from the interplay between health conditions (including mental health and cognitive conditions), contextual characteristics and personal characteristics.2 Supporting the concept that physical health conditions alone do not determine physical function, clinical measures of heart disease severity are not strongly associated with physical function.8 28
Importantly, declines in physical function due to MI may be prevented or ameliorated through individual and system interventions.9 29–34 For example, structured cardiac rehabilitation is effective at preventing or reducing declines in function following an MI.34 The CAPABLE intervention, which incorporates home visits by nurses, occupational therapists and handy workers who can modify the home environment, has shown success in reversing function limitations and slowing decline while also being cost-effective.33 Tailored geriatric care pathways for hospitalised older adults may also prevent or reverse MI-associated declines in function; however, the evidence for these hospital system interventions is mixed and may depend on other aspects of care delivery.32 The prior literature supporting these interventions suggests that the differences in physical function by MI status we observed are not inevitable.
The current study among REGARDS participants and other studies have found that MI was more strongly associated with some measures of function among women, and particularly black women.16 Some prior studies have found stronger associations of MI with function among older compared with younger adults16; among REGARDS participants, we found conflicting evidence across physical function measures. Differences in associations of MI with physical function may result from disparities in receipt of interventions that can prevent or reverse function declines. For example, in the USA, few of those eligible complete cardiac rehabilitation programmes; older individuals, women and black individuals are less likely to receive cardiac rehabilitation, due in part to differential referral.35 Racial disparities in use and quality of home healthcare, which may mitigate function limitations, have also been observed in some but not all studies.36 37
The large geographically diverse study population of black and white American adults in the REGARDS study, multidimensional assessment of physical function and examination of distant and more recent MI are major strengths of this study. However, this population included relatively high-functioning, community-dwelling individuals who agreed to participate in a telephone interview and in-home study visit and completed the function assessments. Non-participation in the second in-home visit or in the gait speed and chair stand assessments because of death or physical function limitations related to MI may have biased results; this concern is somewhat mitigated by sensitivity analyses using inverse probability of attrition weighting that were consistent with the primary analyses. The results may not be generalisable to individuals who are living in residential care settings, those who identify as race other than black or white and to those receiving care outside of the USA. Gait speed measured using an 8-foot timed walk is systematically lower than gait speed measured from longer walking distances.25 The assessment of physical function occurred at varying time points relative to MI events, which could increase heterogeneity within the groups with MI. Among participants with incident MI, the time between the MI event and follow-up assessment had modest associations with PCS but not the other measures of function assessed. Most individuals with MI had a remote history of MI (ie, at least 10 years for those with a history of MI at baseline and no MI event during follow-up; median of 4.7 years for those with MI during follow-up), and results were similar when excluding individuals with MI in the year prior to the function assessment. We adjusted for a number of comorbidities and health behaviours that may increase risk for both MI and physical function limitations, but not all potential shared risk factors were available in this study. We lacked information on MI severity. However, prior literature indicates that measures of cardiac pathology, angina and laboratory exercise capacity are not strongly associated with function among people with coronary heart disease.8 28 38 39 We have limited information on care provided for the MI; therefore, we could not evaluate the role of MI care and disparities in care in our associations, though adjusting for revascularisation did not substantially change our findings. Additionally, it is possible that physical function limitations increase the risk of MI (ie, reverse causation).3
In summary, history of MI was associated with worse physical function across multiple measures, and these associations may be stronger among women, and particularly black women.