Abstract
Background:
Black adults are at high risk for developing heart failure (HF). Physical inactivity and subclinical myocardial injury, assessed by high-sensitivity cardiac troponin-I (hs-cTnI), are independent risk factors for HF.
Objective:
Evaluate the independent associations and interaction between hs-cTnI and physical activity (PA) with risk of HF subtypes, HF with preserved and reduced ejection fraction (HFpEF and HFrEF, respectively).
Methods:
Black adults from the Jackson Heart Study without prevalent HF who had hs-cTnI and self-reported PA assessed at baseline were included. Adjusted Cox models were used to evaluate the independent and joint associations and interaction of hs-cTnI and PA with risk of HFpEF and HFrEF.
Results:
Among 3,959 participants, 25.1% had subclinical myocardial injury (hs-cTnl ≥4 and ≥6 ng/L in women and men, respectively), and 48.2% were inactive (moderate-to-vigorous PA =0 minutes/week). Over 12.0 years of follow-up, 163 and 150 participants had an incident HFpEF and HFrEF event, respectively. In adjusted analysis, higher hs-cTnI (per 1-unit log increase) was associated with higher risk of HFpEF (HR=1.47 [95% CI, 1.25–1.72]) and HFrEF (HR=1.57 [95% CI, 1.35–1.83]). In contrast, higher PA (per 1-unit log increase) was associated with lower risk of HFpEF (HR=0.93 [95% CI, 0.88–0.99]) but not HFrEF. There was a significant interaction between hs-cTnI and PA for risk of HFpEF (p-interaction=0.04) such that inactive participants with subclinical myocardial injury were at higher risk of HFpEF but active participants were not.
Conclusion:
Among Black adults with subclinical myocardial injury, higher levels of PA were associated with attenuated HFpEF risk.
Keywords: Black adults, Heart failure, Physical activity, Subclinical myocardial injury
INTRODUCTION
Heart failure (HF) is more common in Black adults and is associated with a higher burden of downstream hospitalization and risk of morality as compared with other races (1,2). Over the past decade, the rates of incident HF have increased in Black individuals, largely driven by a disproportionately higher incidence of HF with preserved ejection fraction (HFpEF) (3), a subtype of HF without effective disease-modifying therapies (4). In contrast, the incidence of HF with reduced ejection fraction (HFrEF) has remained stable. Thus, novel approaches to risk stratification and prevention of HF, particularly HFpEF, are needed to curb the increasing burden of this epidemic among Black adults.
Recent studies have demonstrated that subclinical myocardial injury, as measured by elevated levels of high-sensitivity cardiac troponin (hs-cTn), may identify a subset of Black individuals at particularly high risk for developing HF (5). However, it is unclear if the risk of HF associated with subclinical myocardial injury is consistent for both HF subtypes, HFpEF and HFrEF. Furthermore, the extent to which the risk of HF outcomes associated with subclinical myocardial injury may be modified by lifestyle factors such as physical activity (PA) and body mass index (BMI) is not known. This is relevant considering the previously reported associations of physical inactivity and obesity with higher burden of subclinical myocardial injury as well as risk of HF, particularly HFpEF (6–8). A better understanding of the interrelationships of PA, BMI and subclinical myocardial injury with risk of HF events is needed to target the highest risk individuals with potentially effective lifestyle interventions.
Accordingly, in this study we evaluated the independent and joint associations of PA, BMI, and subclinical myocardial injury with risk of HFpEF and HFrEF in a Black cohort. We hypothesized that physical inactivity and subclinical myocardial injury were each independently associated with higher risk of both HF subtypes. Based on the stronger association between PA and risk of HFpEF vs. HFrEF (7), we hypothesized that PA would modify the risk of HFpEF associated with subclinical myocardial injury.
METHODS
Study population
The Jackson Heart Study (JHS) is a prospective, community-based, observational cohort study of Black adults living in the Jackson, Mississippi area. Study design, recruitment, and data collection details have been reported previously (9–11). Briefly, Black participants aged 21 to 84 years were recruited to a baseline examination from 2000 to 2004. Participants completed subsequent follow-up visits between 2005 to 2008 (visit 2) and 2009 to 2012 (visit 3). The Institutional Review Boards at participating centers approved the study protocol. All participants provided written informed consent. The present study included participants who had no history of HF prior to 2005 with available data on PA, hs-cTnI, and other relevant clinical covariates at the baseline visit (Supplemental Figure 1).
Clinical variables
During the baseline examination, participants underwent an in-home interview and clinical examination according to standardized protocols (10,11). Details regarding assessment of baseline clinical covariates are described in the Supplemental Methods. Height and weight were measured using standardized protocols. BMI was calculated as weight in kilograms divided by height in meters squared. Obesity was defined as BMI ≥30 kg/m2. As reported previously, brain natriuretic peptide (BNP) was measured using a chemiluminescent immunoassay and Siemens Advia Centaur system (12).
Physical Activity
PA was assessed during the in-home visit as reported previously (13,14). Trained personnel administered the JHS Physical Activity Survey which consisted of 30 items evaluating PA over the past 12 months across the following four domains: 1) work; 2) home, family, yard, and garden; 3) active living; and 4) sports and exercise. To evaluate leisure-time moderate to vigorous PA (MVPA), participants were asked questions regarding the frequency and duration of participation in up to three activities in the sports and exercise domain. The metabolic equivalent of task (MET) was determined for each sport and exercise activity according to the Compendium of Physical Activities (15). Leisure-time MVPA levels were estimated according to the sum of minutes per week performing sport and exercise activities with energy costs ≥3.5 METs as follows: total MET-min/week = moderate activity time * 4.5 + vigorous activity time * 7.5. Participants who performed any (>0 minutes) leisure-time MVPA were considered physically active whereas those who performed no leisure-time MVPA (0 minutes) were defined as physically inactive.
High-Sensitivity Cardiac Troponin I Assessment
Hs-cTnI was measured from thawed plasma samples originally collected during the baseline examination from 2000 to 2004. The ARCHITECT platform (Abbott Diagnostics) was used to measure hs-cTnI and can detect concentrations as low as 1.2 ng/L (16). The coefficient of variation was 10% at a concentration of 3.0 ng/L. Subclinical myocardial injury was defined as hs-cTnI ≥4 and ≥6 ng/L in women and men, respectively, as previously described (5).
Outcomes of Interest
The primary outcomes of interest in this study were incident HFpEF and HFrEF. Incident HF events were defined as the first HF hospitalization event. HF events were formally adjudicated in the JHS starting in January 2005 as previously described (17,18). HF events were further classified as HFpEF (left ventricular ejection fraction [LVEF] ≥50%) or HFrEF (LVEF <50%) according to available cardiac imaging data before and at the time of the initial hospitalization for HF. Overall HF events included all incident HF hospitalizations with and without available LVEF data. The protocol for HF adjudication is described in detail in the Supplemental Methods.
Key secondary outcomes of interest included measures of left ventricular (LV) structure. During visit 3 of the JHS, participants underwent magnetic resonance imaging. A standardized protocol was used to obtain images and measure LV mass and LV end-diastolic volume (LVEDV) (Supplemental Methods).
Statistical Analysis
Participants were categorized into four groups based on PA status and presence or absence of subclinical myocardial injury as follows: 1) active without subclinical myocardial injury; 2) active with subclinical myocardial injury; 3) inactive without subclinical myocardial injury; 4) inactive with subclinical myocardial injury. Baseline characteristics were reported as median (interquartile range) and number (percentage) for continuous and categorical variables, respectively, across categories. Among each PA stratum, continuous variables were compared with Kruskal-Wallis test and categorical variables were compared with chi-square test.
Multivariable-adjusted Cox proportional hazards models were constructed to evaluate the independent associations of continuous measures of PA, BMI, and hs-cTnI with risk of HF events. Separate Cox models were constructed for each HF outcome (HFpEF and HFrEF) with censoring for death, end of study follow-up, and the other HF subtype. Due to skewed distributions, PA (total MET-min/week + 1) and hs-cTnI levels were log transformed. Models included the following covariates selected a priori based on biologic plausibility and prior studies (5,7): demographics (age, sex), traditional cardiovascular disease (CVD) risk factors (BMI, systolic blood pressure, smoking status, history of hypertension, history of diabetes), history of CVD, laboratory values (HbA1c, estimated GFR, hs-cTnI), and PA level. Multiplicative interaction terms were included in the adjusted Cox models to evaluate whether PA or BMI modified the association between hs-cTnI and risk of HF events (PA * hs-cTnI; BMI * hs-cTnI).
Cumulative incidence plots were used to evaluate the risk of incident HF events across categories stratified by PA and subclinical myocardial injury status. Separate Cox models were created to evaluate the independent associations between these categories and risk of HFpEF and HFrEF with adjustment for the covariates listed above (except PA and hs-cTnI). Sensitivity analyses were also performed additionally adjusting for BNP. Furthermore, in the subset of participants who had available cardiac MRI data on visit 3 and did not develop HF prior to cardiac MRI assessment, the association between baseline PA / subclinical myocardial injury status and measures of LV mass and LVEDV were assessed using multivariable-adjusted linear regression models with adjustment for the same potential confounders listed previously.
The interrelationships of obesity and subclinical myocardial injury with risk of HF were evaluated in similar analyses. Participants were stratified according to obesity and subclinical myocardial injury status into the following four groups: 1) non-obese without subclinical myocardial injury; 2) non-obese with subclinical myocardial injury; 3) obese without subclinical myocardial injury; 4) obese with subclinical myocardial injury. The associations of obesity / subclinical myocardial injury categories with risk of HF events were assessed using adjusted Cox models with inclusion of similar covariates as described previously.
All statistical analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, North Carolina, USA). Two-sided p-values <0.05 were considered statistically significant.
RESULTS
Baseline characteristics
Among the 3,959 participants (63.2% women) included in the present study, 48.2% were physically inactive. Prevalence of subclinical myocardial injury in the overall cohort was 25.1% with higher proportions among participants who were inactive vs. active (29.1% vs. 21.3%, respectively [p-value <0.001]). Baseline characteristics of the study participants according to PA status stratified by presence of subclinical myocardial injury are shown in Table 1. Across PA categories, participants with subclinical myocardial injury were older, had a greater burden of CVD risk factors, including higher BMI, and more commonly used antihypertensive and glucose lowering medications. Participants with subclinical myocardial injury who were inactive had the greatest burden of traditional CVD risk factors.
Table 1.
Baseline characteristics stratified by physical activity and subclinical myocardial injury status.
| Physically active (n = 2,051) | Physically inactive (n = 1,908) | |||||
|---|---|---|---|---|---|---|
| No subclinical myocardial injury (n = 1,614) | Subclinical myocardial injury (n = 437) | P value | No subclinical myocardial injury (n = 1,352) | Subclinical myocardial injury (n = 556) | P value | |
| Age, years | 50.6 (42.3, 59.9) | 59.9 (50.3, 67.4) | <0.001 | 55.7 (46.0, 64.1) | 64.1 (56.1, 71.5) | <0.001 |
| Male, % | 631 (39.1) | 158 (36.2) | 0.26 | 492 (36.4) | 175 (31.5) | 0.04 |
| Weight, kg | 87.0 (75.3, 100) | 89.4 (78.0, 104.0) | 0.01 | 87.5 (75.4, 102.4) | 89.0 (77.5, 104.0) | 0.12 |
| Body mass index, kg/m2 | 30.0 (26.4, 34.4) | 30.7 (27.7, 35.5) | 0.001 | 30.5 (26.6, 35.7) | 31.8 (27.8, 36.5) | 0.002 |
| Hypertension, % | 706 (43.7) | 310 (70.9) | <0.001 | 703 (52.0) | 447 (80.4) | <0.001 |
| Systolic BP, mmHg | 122 (114, 132) | 130 (121, 143) | <0.001 | 125 (116, 135) | 134 (123, 147) | <0.001 |
| Diastolic BP, mmHg | 75 (70, 81) | 77 (71, 83) | 0.03 | 76 (70, 82) | 76 (68, 82) | 0.58 |
| Antihypertensive medication, % | 619 (38.9) | 277 (64.3) | <0.001 | 578 (43.4) | 385 (70.9) | <0.001 |
| Diabetes, % | 231 (14.3) | 110 (25.2) | <0.001 | 275 (20.3) | 182 (32.7) | <0.001 |
| Hemoglobin A1c, % | 5.5 (5.2, 6.0) | 5.7 (5.3, 6.3) | <0.001 | 5.7 (5.3, 6.1) | 5.9 (5.5, 6.5) | <0.001 |
| Diabetes medication, % | 149 (9.8) | 70 (17.3) | <0.001 | 165 (13.1) | 124 (24.9) | <0.001 |
| Current smoker, % | 151 (9.5) | 33 (7.6) | 0.23 | 222 (16.5) | 71 (12.9) | 0.046 |
| History of cardiovascular disease, % | 70 (4.3) | 56 (12.8) | <0.001 | 96 (7.1) | 83 (14.9) | <0.001 |
| History of coronary heart disease, % | 42 (2.6) | 39 (8.9) | <0.001 | 60 (4.4) | 55 (9.9) | <0.001 |
| History of myocardial infarction, % | 24 (1.5) | 22 (5.0) | <0.001 | 45 (3.3) | 32 (5.8) | 0.01 |
| Total cholesterol, mg/dL | 195 (171, 221) | 203 (179, 224) | 0.003 | 196 (171, 223) | 199 (175, 228) | 0.08 |
| HDL, mg/dL | 50 (41, 59) | 50 (42, 61) | 0.15 | 49 (41, 60) | 50 (41, 61) | 0.28 |
| LDL, mg/dL | 124 (101, 147) | 128 (106, 149) | 0.06 | 125 (102, 150) | 125 (101, 153) | 0.75 |
| Triglycerides, mg/dL | 86 (61, 123) | 94 (70, 132) | 0.004 | 88 (66, 121) | 96 (69, 136) | 0.004 |
| eGFR, mL/min per 1.73 m2 | 99.0 (85.6, 112.1) | 90.5 (77.5, 103.6) | <0.001 | 97.8 (84.5, 110.3) | 87.0 (69.0, 101.2) | <0.001 |
| Left ventricular hypertrophy, % | 47 (2.9) | 45 (10.3) | <0.001 | 49 (3.6) | 74 (13.3) | <0.001 |
| LV mass, g | 135.2 (115.2, 159.9) | 159.2 (130.4, 191.5) | <0.001 | 138.4 (118.8, 158.6) | 160.4 (132.8, 195.1) | <0.001 |
| hs-cTnI, ng/L | 2.5 (1.9, 3.2) | 7.6 (5.2, 13.8) | <0.001 | 2.7 (2.0, 3.4) | 7.4 (5.4, 13.2) | <0.001 |
| BNP, pg/mL | 5.7 (2.0, 13.1) | 10.4 (4.0, 23.9) | <0.001 | 6.7 (2.3, 14.3) | 14.3 (5.5, 30.5) | <0.001 |
| Physical activity, MET-min/week | 467 (207, 1096) | 452 (197, 1055) | 0.94 | 0 (0, 0) | 0 (0, 0) | >0.99 |
Physically inactive refers to 0 minutes/week of MVPA. Subclinical myocardial injury refers to hs-cTnI ≥6 ng/L in men and ≥4 ng/L in women. Data presented as median (inter-quartile range) or %. Comparison performed using χ2 for categorical variables and Kruskal–Wallis for continuous variables.
Abbreviations: BP, blood pressure; BNP, brain natriuretic peptide; eGFR, estimated glomerular filtration rate; hs-cTnI: high-sensitivity cardiac troponin I; LDL, low-density lipoprotein; LV, left ventricular; MVPA, moderate to vigorous physical activity.
Physical activity, subclinical myocardial injury, and risk of heart failure
Over a median follow-up of 12.0 (interquartile range 10.6–12.3) years, there were 322 incident HF events. There were 163 (50.6%) and 150 (46.6%) incident HFpEF and HFrEF events, respectively, with the remaining HF events (2.8%) missing LVEF data. In multivariable-adjusted analysis, greater subclinical myocardial injury was independently associated with higher risk of both HF subtypes (HFpEF: hazard ratio [HR] per 1-unit higher log[hs-cTnI], 1.47 [95% CI, 1.25–1.72]; HFrEF: HR per 1 unit higher log[hs-cTnI], 1.57 [95% CI, 1.35–1.83]) (Table 2). In contrast, PA was significantly associated with lower risk of HFpEF (HR per 1 unit higher log[MET-min/week + 1], 0.93 [95% CI, 0.88–0.99]) but not HFrEF. A statistically significant interaction was observed between PA and hs-cTnI for risk of HFpEF (p-interaction=0.04) but not HFrEF.
Table 2.
Multivariable-adjusted associations of physical activity, subclinical myocardial injury, and other clinical characteristics with risk of HFpEF and HFrEF.
| HFpEF | HFrEF | |||
|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | |
| Log PA (per 1 unit increase log[MET-min/week + 1]) | 0.93 (0.88, 0.99) | 0.01 | 0.96 (0.91, 1.02) | 0.17 |
| Log hs-cTnI (per 1-unit increase log[hs-cTnI]) | 1.47 (1.25, 1.72) | <0.001 | 1.57 (1.35, 1.83) | <0.001 |
| Other clinical characteristics | ||||
| Age (per 1-year increase) | 1.07 (1.05, 1.09) | <0.001 | 1.05 (1.03, 1.07) | <0.001 |
| Male (vs. female) | 0.92 (0.64, 1.31) | 0.64 | 1.40 (0.99, 1.97) | 0.05 |
| BMI (per 1 kg/m2 increase) | 1.05 (1.03, 1.08) | <0.001 | 0.99 (0.97, 1.02) | 0.79 |
| Hypertension (vs. no hypertension) | 1.77 (1.12, 2.80) | 0.02 | 1.18 (0.77, 1.79) | 0.45 |
| Diabetes (vs. no diabetes) | 1.59 (1.06, 2.40) | 0.03 | 1.44 (0.92, 2.27) | 0.11 |
| Systolic BP (per 1 mm Hg increase) | 1.00 (0.99, 1.01) | 0.95 | 1.00 (0.99, 1.01) | 0.45 |
| HbAlc (per 1% increase) | 1.15 (1.01, 1.30) | 0.04 | 1.15 (1.01, 1.31) | 0.04 |
| CVD history (vs. no CVD history) | 1.65 (1.12, 2.42) | 0.01 | 1.43 (0.93, 2.21) | 0.10 |
| eGFR (per 1 mL/min per 1.73 m2 increase) | 0.98 (0.98, 0.99) | <0.001 | 0.99 (0.98, 1.00) | 0.11 |
| Current smoker (vs. no current smoker) | 2.53 (1.60, 3.99) | <0.001 | 1.56 (0.97, 2.53) | 0.07 |
Model is adjusted for the following covariates: age, sex, body mass index, history of hypertension, history of diabetes, systolic BP, HbA1c, history of CVD, eGFR, smoking status, log hs-cTnI, log PA.
Abbreviations: BMI, body mass index; BP, blood pressure; CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate; HbA1c, hemoglobin A1c; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; hs-cTnI, high-sensitivity cardiac troponin I; PA, physical activity.
The cumulative incidence of HF subtypes across categories stratified by PA and subclinical myocardial injury status are shown in Figure 1. In multivariable adjusted analysis, the risk of HFpEF associated with subclinical myocardial injury differed according to PA status such that higher risk of HFpEF was observed among participants with subclinical myocardial injury who were inactive (HR, 2.10 [95% CI, 1.33–3.30]; referent group, active, no injury) but not those who were active (HR, 0.95 [95% CI, 0.53, 1.70]; referent group, active, no injury) (Table 3). In contrast, the risk of HFrEF was consistently higher among individuals with subclinical myocardial injury at baseline irrespective of their PA status. Compared with active participants who had no subclinical myocardial injury (referent group), HFrEF risk was higher both among those with subclinical myocardial injury who were inactive (HR, 2.51 [95% CI, 1.56–4.06]) and active (HR, 1.97 [95% CI, 1.16–3.33]).
Figure 1. Risk of hospitalization for HFpEF (top) and HFrEF (bottom) among participants stratified by physical activity and subclinical myocardial injury status.
Physically inactive refers to 0 minutes/week of MVPA. Subclinical myocardial injury refers to hs-cTnI ≥6 ng/L in men and ≥4 ng/L in women.
Abbreviations: HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fractions; hs-cTnI = high-sensitivity cardiac troponin I; MVPA = moderate to vigorous physical activity.
Table 3.
Multivariable-adjusted associations between categorical measures of physical activity and subclinical myocardial injury with risk of HFpEF and HFrEF.
| HFpEF | HFrEF | |||
|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | |
| Referent group: active, no subclinical myocardial injury | ||||
| Active, subclinical myocardial injury | 0.95 (0.53, 1.70) | 0.85 | 1.97 (1.16, 3.33) | 0.01 |
| Inactive, no subclinical myocardial injury | 1.09 (0.69, 1.72) | 0.71 | 1.11 (0.69, 1.77) | 0.67 |
| Inactive, subclinical myocardial injury | 2.10 (1.33, 3.30) | 0.001 | 2.51 (1.56, 4.06) | <0.001 |
Physically inactive refers to 0 minutes/week of MVPA. Subclinical myocardial injury refers to hs-cTnI ≥6 ng/L in men and ≥4 ng/L in women. Model is adjusted for the following covariates: age, sex, body mass index, history of hypertension, history of diabetes, systolic BP, HbA1c, history of CVD, eGFR, smoking status, PA/injury groups.
Abbreviations: BP, blood pressure; CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate; HbA1c, glycated hemoglobin; HF, heart failure; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; MVPA, moderate to vigorous physical activity; PA, physical activity.
In sensitivity analysis, among participants with available BNP data at baseline (n=3,170), the association of subclinical myocardial injury with risk of HFrEF was modestly attenuated after adjusting for BNP (Supplemental Table 1). However, the overall association and relative differences in risk of HF outcomes across categories of PA / subclinical myocardial injury status with risk of HFpEF and HFrEF were comparable to that observed in the primary analysis (Supplemental Table 1). The interaction between PA and hs-cTnI for risk of HFpEF remained statistically significant after additional adjustment for BNP (p-interaction=0.04). In the adjusted models, BNP was significantly associated with risk of HFpEF (HR per 1-unit higher log[BNP], 1.37 [95% CI 1.17–1.61]) and HFrEF (HR per 1-unit higher log[BNP], 1.27 [95% CI 1.07–1.51]).
Physical activity, subclinical myocardial injury, and cardiac structure & function on f/u
The association of PA and subclinical myocardial injury with measures of LV mass and LVEDV on follow-up was assessed in a subset of participants who did not develop HF prior to undergoing cardiac magnetic resonance imaging during visit 3 (Supplemental Table 2). Subclinical myocardial injury was associated with higher LV mass among both active as well as inactive individuals (Std.ß: 13.0 and 15.0 respectively, p-value <0.001 for both, ref.: Active with no injury). In contrast, the association between subclinical myocardial injury and LVEDV on follow-up differed by baseline PA status. Subclinical myocardial injury at baseline was significantly associated with higher LVEDV on follow-up only among inactive (Std.ß: 8.78, p-value = 0.03) but not active individuals (Std.ß: 2.73, p-value = 0.48). Among individuals without subclinical myocardial injury, those who were inactive had significantly lower LVEDV on follow-up as compared with active individuals (Std.ß −3.79, p-value = 0.049).
Obesity, subclinical myocardial injury, and risk of heart failure
Obesity had differential associations with HF subtypes such that higher BMI was significantly associated with higher risk of HFpEF (HR per 1 kg/m2 higher BMI, 1.05 [95% CI 1.03–1.08]) but not HFrEF (Table 2). Compared to non-obese individuals without subclinical myocardial injury (ref. group), obese individuals had higher risk of HFpEF irrespective of subclinical myocardial injury status (Table 4). Furthermore, subclinical myocardial injury in absence of obesity was not associated with higher risk of HFpEF. In contrast, the risk of HFrEF was only elevated among individuals with subclinical myocardial injury irrespective of obesity status with the highest risk noted in those with obesity (HR, 2.20 [95% CI, 1.38, 3.51]; referent group, non-obese, no injury). Additional adjustment for BNP modestly attenuated the association of obesity with risk of HFpEF but the overall pattern of association was mostly similar to the primary analysis (Supplemental Table 3).
Table 4.
Multivariable-adjusted associations between categorical measures of obesity and subclinical myocardial injury with risk of HFpEF and HFrEF.
| HFpEF | HFrEF | |||
|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | |
| Referent group: Non-obese, no subclinical myocardial injury | ||||
| Non-obese, subclinical myocardial injury | 1.42 (0.82, 2.47) | 0.21 | 1.87 (1.14, 3.06) | 0.01 |
| Obese, no subclinical myocardial injury | 1.65 (1.03, 2.64) | 0.04 | 0.93 (0.58, 1.50) | 0.78 |
| Obese, subclinical myocardial injury | 2.79 (1.74, 4.46) | <0.001 | 2.20 (1.38, 3.51) | <0.001 |
Obesity refers to BMI ≥30 kg/m2. Subclinical myocardial injury refers to hs-cTnI ≥6 ng/L in men and ≥4 ng/L in women. Model is adjusted for the following covariates: age, sex, history of hypertension, history of diabetes, systolic BP, HbA1c, history of CVD, eGFR, smoking status, log PA, obesity/injury groups.
Abbreviations: BMI, body mass index; BP, blood pressure; CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate; HbA1c, glycated hemoglobin; HF, heart failure; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; PA, physical activity.
DISCUSSION
Among Black adults, we observed that subclinical myocardial injury was independently associated with higher risk of both HF subtypes, HFpEF and HFrEF. In contrast, self-reported PA was inversely associated with risk of HFpEF, but not HFrEF, after accounting for traditional CVD risk factors. Additionally, PA modified the association between subclinical myocardial injury and risk of HFpEF such that active individuals with subclinical myocardial injury had comparable risk to those who had no subclinical myocardial injury (Central Illustration). Conversely, risk of HFrEF was associated with subclinical myocardial injury irrespective of PA levels.
Central Illustration. Physical activity modifies the association between subclinical myocardial injury and risk of HFpEF but not HFrEF.
Physically inactive refers to 0 minutes/week of MVPA. Subclinical myocardial injury refers to hs-cTnI ≥6 ng/L in men and ≥4 ng/L in women.
Abbreviations same as figure 1.
Previous studies have demonstrated a dose-dependent, inverse association between higher levels of PA and lower risk of HF (19,20). Furthermore, prior studies have reported a differential association between PA and risk of HF subtypes with a stronger association between physical inactivity and risk of HFpEF vs. HFrEF (7). The findings from the present study add to the existing literature by confirming the stronger association between physical inactivity and risk of HFpEF compared with HFrEF in a cohort of community-dwelling Black adults.
Subclinical myocardial injury is one of the strongest predictors of HF risk in black adults. Prior work from the JHS demonstrated a 2- to 3-fold higher risk of HF among individuals with subclinical myocardial injury (5). In the present study, we extend these observations and demonstrated that the risk of HF associated with subclinical myocardial injury is consistent for both HFpEF and HFrEF. Our study findings differ from the observation from a prior pooled cohort study that demonstrated a stronger association between subclinical myocardial injury and risk of HFrEF vs. HFpEF (21). Several factors may underlie the observed differences across the two studies. First, there was an 8-fold difference in the proportion of Black adults enrolled in the two studies (100% [3,959/3,959] vs. 12% [2,681/22,756]). Black adults have a greater burden of subclinical myocardial injury compared with other racial or ethnic groups which may reflect differences in underlying pathophysiology, cardiac structure and function, or burden of accompanying risk factors such as uncontrolled hypertension (8). Second, while the present study used levels of hs-cTnI to assess the burden of subclinical myocardial injury, study cohorts with hs-cTnI and hs-cTnT based assessments of subclinical myocardial injury were pooled together in the prior study (21). Recent studies have demonstrated significant differences regarding pathophysiological determinants of hs-cTnI and hs-cTnT, as well as in the predictive ability for HF (22). Third, the proportion of HF cases that were able to be adjudicated as HFpEF vs. HFrEF differs across cohorts, which is another factor that may underlie the observed variability in association between subclinical myocardial injury and risk of HF subtypes. For example, in our study, HF subtype could not be determined in 3% [9/322] of cases as compared with 10% [11/114] in the Atherosclerosis Risk in Communities Study (23) and 30% in the pooled cohort analysis [621/2,095] (21). Finally, unlike the present study, de Boer, et al. did not account for PA levels in their adjusted analysis examining the association of subclinical myocardial injury with risk of HF subtypes.
A novel finding of the present study is the significant interaction between PA and subclinical myocardial injury for the risk of HFpEF among Black adults. We observed that higher levels of PA favorably modified the risk of HFpEF associated with subclinical myocardial injury, highlighting the potential role of PA interventions in modifying HFpEF risk in high-risk individuals. The mechanisms through which PA may modify hs-cTnI associated risk of HFpEF are not well-known. Prior studies have demonstrated that higher levels of baseline PA have lower probability of worsening injury burden on follow-up (24). Furthermore, PA interventions have been shown to attenuate subclinical myocardial injury burden over time (25). Thus, it is plausible that higher PA may lower the risk of HFpEF by preventing downstream worsening of subclinical myocardial injury. Prior studies have also demonstrated that higher cardiorespiratory fitness at baseline and less decline in fitness over time were each significantly associated with lower risk of HFpEF (26). Higher PA may also identify individuals who engage in more routine exercise behavior and thus may have lower decline in cardiorespiratory fitness and less weight gain with aging, which may contribute to lower downstream risk of HFpEF (27). Future studies are needed to determine if exercise training interventions may modify risk of HFpEF among high risk individuals with subclinical myocardial injury.
We also observed different LV remodeling patterns on follow-up among individuals based on their baseline PA and subclinical myocardial injury status. Inactive individuals without subclinical myocardial injury at baseline had significantly lower LV volume, an intermediate phenotype associated with greater risk of HFpEF (28). Furthermore, the association of subclinical myocardial injury with LV volume differed according to baseline PA status. Subclinical myocardial injury at baseline was associated with significantly higher LV volume on follow-up only among inactive but not active individuals. These findings provide insights into the potential mechanisms underlying the interaction between PA and hs-TnI levels for risk of HFpEF.
In the present study, we evaluated the interrelationships of obesity and subclinical myocardial injury with risk of HF. Consistent with prior studies, obesity was differentially associated with risk of HF subtypes (7,29). We extend previous study findings and demonstrate that obesity was significantly associated with risk of HFpEF independent of subclinical myocardial injury, but not HFrEF, in a cohort of Black adults. Obesity appears to be an important underlying factor for HFpEF risk irrespective of subclinical myocardial injury status. In contrast, subclinical myocardial injury appears to be closely linked with HFrEF risk independent of measures of obesity. Taken together, there are distinct risk factors and underlying pathophysiologic drivers of risk for HFpEF and HFrEF. Our study findings are consistent with prior studies that suggest peripheral factors are more important determinants of HFpEF whereas direct cardiac impairment underlies the development of HFrEF (4,29).
Strengths and limitations
The strengths of the present study include the large sample size of Black adults with inclusion of approximately two-thirds women, objective assessment of subclinical myocardial injury assessed by hs-cTnI, long study follow-up, and standardized assessment of HF events with classification of HF subtypes in the majority of cases. There are also noteworthy limitations. First, PA was self-reported and is susceptible to recall bias and misclassification. However, the JHS Physical Activity Survey has high reproducibility (intra-class correlation = 0.99) and has been validated with accelerometer (rho = 0.24) and pedometer (rho = 0.32) data (30). Furthermore, discrete PA categories (0 vs. >0 minutes/week of MVPA) may be less susceptible to misclassification. Additionally, the present study focused on MVPA assessed from the sports and exercise domain which may underestimate PA but this would be expected to bias the study findings towards the null. Second, there is a potential for reverse causation as physical inactivity and subclinical myocardial injury may be markers of prevalent HF. However, baseline assessment of covariates started as early as 2000 and adjudication of HF events did not start until after 2005 suggesting a lower likelihood of reverse causation. Third, cardiorespiratory fitness and muscle strength may explain in part the association between PA and risk of HFpEF but these data were not available for inclusion in the adjusted models (31). Fourth, incident HF was based on the first HF hospitalization, data regarding LV mass at the time of the incident hospitalization were not routinely captured, and adjudication did not require biomarker abnormalities which may lead to misclassification of HF events, especially HFpEF. However, HF events were adjudicated using standardized protocols and nearly all HF events were classified into subtypes using objective LVEF cutoffs. Fifth, the present study design is observational, and, thus, residual or unmeasured confounding cannot be entirely excluded. Finally, our findings should be interpreted in the context of the Black study population and may not be generalizable to other racial or ethnic groups.
CONCLUSIONS
Among community-dwelling Black adults, physical inactivity and subclinical myocardial injury are each independently associated with higher risk of HFpEF. Physically active adults with subclinical myocardial injury have similar risk of HFpEF as those who were active with no subclinical myocardial injury. Future studies evaluating the impact of PA on longitudinal changes in subclinical myocardial injury are needed to evaluate whether targeted lifestyle interventions can prevent HFpEF.
Supplementary Material
Clinical Perspectives:
Competency in Medical Knowledge:
Among Black adults from the Jackson Heart Study, the combination of physical inactivity and subclinical myocardial injury was associated with a 2-fold higher risk of heart failure with preserved ejection fraction compared with those who were active without evidence of subclinical myocardial injury.
Translational Outlook:
Future studies are needed to evaluate whether a lifestyle intervention focused on physical activity can lower the burden of subclinical myocardial injury and its associated risk of heart failure with preserved ejection fraction among high-risk individuals.
Acknowledgements:
The authors thank the study participants, staff, and investigators of the Jackson Heart Study. The authors thank Rachael Whitehead of Houston Methodist Research Institute for helping create the central illustration.
Funding & Disclosures:
Dr. Neeland reports consulting fees from Boehringer Ingelheim/Lilly Alliance and Merck.
Dr. deFilippi reports grant support from Roche Diagnostics, Ortho Diagnostics, and Siemens Healthineers, consulting fees from Roche Diagnostics, Abbott Diagnostics, Ortho Clinical Diagnostics, Quidel, FujiRebio and Siemen’s Health Care Diagnostics. Dr. deFilippi receives royalty payments from UpToDate. Dr. deFilippi has been named a co-owner on a patent awarded to the University of Maryland (US Patent Application Number: 15/309,754) entitled: “Methods for Assessing Differential Risk for Developing Heart Failure.” Dr. deFilippi receives funding from the National Center for Advancing Translational Science of the National Institutes of Health Award UL1TR003015.
Dr. Seliger receives funding from Roche Diagnostics.
Dr. de Lemos reports grant support from Roche Diagnostics and Abbott Diagnostics, consulting fees from Roche Diagnostics, Abbott Diagnostics, Ortho Clinical Diagnostics, Quidel Cardiovascular, Inc, and Siemen’s Health Care Diagnostics. Dr. de Lemos has been named a co-owner on a patent awarded to the University of Maryland (US Patent Application Number: 15/309,754) entitled: “Methods for Assessing Differential Risk for Developing Heart Failure.”
Dr. Mentz received research support and honoraria from Abbott, American Regent, Amgen, AstraZeneca, Bayer, Boehringer Ingelheim/Eli Lilly, Boston Scientific, Cytokinetics, Fast BioMedical, Gilead, Innolife, Medtronic, Merck, Novartis, Relypsa, Respicardia, Roche, Sanofi, Vifor, and Windtree Therapeutics.
Dr. Berry received funding from grant 14SFRN20740000 from the American Heart Association prevention network and salary support from Abbott Diagnostics.
This project was funded by the Strategically Focused Research Network Grant for Prevention from the American Heart Association to University of Texas Southwestern Medical Center, Dallas, and Northwestern University School of Medicine, Chicago.
Dr. Pandey has served on the advisory board of Roche Diagnostics, has received non-financial support from Pfizer and Merck, has received research support from Texas Health Resources Clinical Scholarship, the Gilead Sciences Research Scholar Program, the National Institute of Aging GEMSSTAR Grant (1R03AG067960-01), and Applied Therapeutics.
The funding for biomarker assays was provided by Abbott Diagnostics.
Abbreviations:
- HF
heart failure
- HFpEF
heart failure with preserved ejection fraction
- HFrEF
heart failure with reduced ejection fraction
- hs-cTnI
high-sensitivity cardiac troponin I
- JHS
Jackson Heart Study
- MVPA
moderate to vigorous physical activity
- PA
physical activity
REFERENCES
- 1.Brown DW, Haldeman GA, Croft JB, Giles WH, Mensah GA. Racial or ethnic differences in hospitalization for heart failure among elderly adults: Medicare, 1990 to 2000. Am Heart J 2005;150:448–54. [DOI] [PubMed] [Google Scholar]
- 2.Sidney S, Go AS, Jaffe MG, Solomon MD, Ambrosy AP, Rana JS. Association Between Aging of the US Population and Heart Disease Mortality From 2011 to 2017. JAMA Cardiol 2019;4:1280–1286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Chang PP, Wruck LM, Shahar E et al. Trends in Hospitalizations and Survival of Acute Decompensated Heart Failure in Four US Communities (2005–2014): ARIC Study Community Surveillance. Circulation 2018;138:12–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Pandey A, Patel KV, Vaduganathan M et al. Physical Activity, Fitness, and Obesity in Heart Failure With Preserved Ejection Fraction. JACC Heart Fail 2018;6:975–982. [DOI] [PubMed] [Google Scholar]
- 5.Pandey A, Keshvani N, Ayers C et al. Association of Cardiac Injury and Malignant Left Ventricular Hypertrophy With Risk of Heart Failure in African Americans: The Jackson Heart Study. JAMA Cardiol 2019;4:51–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Florido R, Ndumele CE, Kwak L et al. Physical Activity, Obesity, and Subclinical Myocardial Damage. JACC Heart Fail 2017;5:377–384. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Pandey A, LaMonte M, Klein L et al. Relationship Between Physical Activity, Body Mass Index, and Risk of Heart Failure. J Am Coll Cardiol 2017;69:1129–1142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.de Lemos JA, Drazner MH, Omland T et al. Association of troponin T detected with a highly sensitive assay and cardiac structure and mortality risk in the general population. JAMA 2010;304:2503–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Taylor HA Jr. The Jackson Heart Study: an overview. Ethn Dis 2005;15:S6–1-3. [PubMed] [Google Scholar]
- 10.Taylor HA Jr., Wilson JG, Jones DW et al. Toward resolution of cardiovascular health disparities in African Americans: design and methods of the Jackson Heart Study. Ethn Dis 2005;15:S6–4-17. [PubMed] [Google Scholar]
- 11.Fuqua SR, Wyatt SB, Andrew ME et al. Recruiting African-American research participation in the Jackson Heart Study: methods, response rates, and sample description. Ethn Dis 2005;15:S6–18-29. [PubMed] [Google Scholar]
- 12.Fox ER, Musani SK, Bidulescu A et al. Relation of obesity to circulating B-type natriuretic peptide concentrations in blacks: the Jackson Heart Study. Circulation 2011;124:1021–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Dubbert PM, Carithers T, Ainsworth BE, Taylor HA Jr., Wilson G, Wyatt SB. Physical activity assessment methods in the Jackson Heart Study. Ethn Dis 2005;15:S6–56-61. [PubMed] [Google Scholar]
- 14.Diaz KM, Booth JN, 3rd, Seals SR et al. Sedentary behavior and subclinical atherosclerosis in African Americans: cross-sectional analysis of the Jackson heart study. Int J Behav Nutr Phys Act 2016;13:31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Ainsworth BE, Haskell WL, Whitt MC et al. Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc 2000;32:S498–504. [DOI] [PubMed] [Google Scholar]
- 16.Apple FS, Collinson PO, Biomarkers ITFoCAoC. Analytical characteristics of high-sensitivity cardiac troponin assays. Clin Chem 2012;58:54–61. [DOI] [PubMed] [Google Scholar]
- 17.Keku E, Rosamond W, Taylor HA Jr. et al. Cardiovascular disease event classification in the Jackson Heart Study: methods and procedures. Ethn Dis 2005;15:S6–62-70. [PubMed] [Google Scholar]
- 18.Kamimura D, Cain LR, Mentz RJ et al. Cigarette Smoking and Incident Heart Failure: Insights From the Jackson Heart Study. Circulation 2018;137:2572–2582. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Pandey A, Garg S, Khunger M et al. Dose-Response Relationship Between Physical Activity and Risk of Heart Failure: A Meta-Analysis. Circulation 2015;132:1786–94. [DOI] [PubMed] [Google Scholar]
- 20.Koo P, Gjelsvik A, Choudhary G et al. Prospective Association of Physical Activity and Heart Failure Hospitalizations Among Black Adults With Normal Ejection Fraction: The Jackson Heart Study. J Am Heart Assoc 2017;6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.de Boer RA, Nayor M, deFilippi CR et al. Association of Cardiovascular Biomarkers With Incident Heart Failure With Preserved and Reduced Ejection Fraction. JAMA Cardiol 2018;3:215–224. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Jia X, Sun W, Hoogeveen RC et al. High-Sensitivity Troponin I and Incident Coronary Events, Stroke, Heart Failure Hospitalization, and Mortality in the ARIC Study. Circulation 2019;139:2642–2653. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Myhre PL, Claggett B, Ballantyne CM et al. Association Between Circulating Troponin Concentrations, Left Ventricular Systolic and Diastolic Functions, and Incident Heart Failure in Older Adults. JAMA Cardiol 2019;4:997–1006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.deFilippi CR, de Lemos JA, Tkaczuk AT et al. Physical activity, change in biomarkers of myocardial stress and injury, and subsequent heart failure risk in older adults. J Am Coll Cardiol 2012;60:2539–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.deFilippi CR, de Lemos JA, Newman AB et al. Impact of moderate physical activity on the longitudinal trajectory of a cardiac specific biomarker of injury: Results from a randomized pilot study of exercise intervention. Am Heart J 2016;179:151–6. [DOI] [PubMed] [Google Scholar]
- 26.Pandey A, Patel KV, Bahnson JL et al. Association of Intensive Lifestyle Intervention, Fitness, and Body Mass Index With Risk of Heart Failure in Overweight or Obese Adults With Type 2 Diabetes Mellitus: An Analysis From the Look AHEAD Trial. Circulation 2020;141:1295–1306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Elagizi A, Carbone S, Lavie CJ, Mehra MR, Ventura HO. Implications of obesity across the heart failure continuum. Prog Cardiovasc Dis 2020;63:561–569. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Velagaleti RS, Gona P, Pencina MJ et al. Left ventricular hypertrophy patterns and incidence of heart failure with preserved versus reduced ejection fraction. Am J Cardiol 2014;113:117–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Ho JE, Lyass A, Lee DS et al. Predictors of new-onset heart failure: differences in preserved versus reduced ejection fraction. Circ Heart Fail 2013;6:279–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Smitherman TA, Dubbert PM, Grothe KB et al. Validation of the Jackson Heart Study Physical Activity Survey in African Americans. J Phys Act Health 2009;6 Suppl 1:S124–32. [DOI] [PubMed] [Google Scholar]
- 31.Carbone S, Kirkman DL, Garten RS et al. Muscular Strength and Cardiovascular Disease: AN UPDATED STATE-OF-THE-ART NARRATIVE REVIEW. J Cardiopulm Rehabil Prev 2020;40:302–309. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.