Association of Diastolic Dysfunction with Reduced Cardiorespiratory Fitness in Adults Living with HIV

Abstract

Despite the high prevalence of diastolic dysfunction in adults living with HIV, the impact on cardiorespiratory fitness (CRF) is understudied. The objective of this cross-sectional study was to investigate the relationship between cardiac function and CRF in adults with HIV. Adults receiving antiretroviral therapy with no history of coronary artery disease (CAD) or heart failure were eligible to participate. Cardiac function was assessed by resting Doppler echocardiography. CRF was measured by oxygen utilization at peak exercise (VO₂peak). The majority of participants were African American (86%) and male (97%) with a mean [standard deviation (SD)] age of 56.6 (7.1) years and median CD4 lymphocyte count of 492 cells/mL. The mean (SD) VO₂peak was 26.1 (5.5) mL/(kg·min). Age, diabetes, hypertension, and hemoglobin were associated with VO₂peak. Overall, diastolic dysfunction was present in 38% and was associated with lower VO₂peak (p < 0.05). VO₂peak was lower among those with impaired myocardial relaxation (e’ <8 cm/s) compared with normal relaxation [mean ± SE mL/(kg·min), 25.2 ± 0.6 vs. 27.7 ± 0.9, p < 0.05]. Adjusted for age and clinical factors, each unit increase in left ventricular relaxation (E/A) was associated with an average 4.4 mL/(kg·min) higher VO₂peak, representing more than one metabolic equivalent. We conclude that diastolic dysfunction is independently associated with clinically significant low CRF in adults with HIV and no history of CAD or heart failure. These results highlight the importance of recognizing diastolic dysfunction in individuals living with HIV regardless of their cardiovascular disease history.

Introduction

The increased risk of cardiovascular disease (CVD) in adults living with HIV is well characterized and independent of traditional risk factors.¹ In addition to coronary artery disease (CAD), adults with HIV have an increased risk of heart failure, especially heart failure with preserved ejection fraction (HFpEF), which is driven by diastolic dysfunction.^2–5 Diastolic dysfunction has been reported in 37%² of asymptomatic adults living with HIV with an incidence rate of 8.2/100 person-years over 4 years.⁶ Mechanisms mediating cardiac dysfunction in the setting of HIV include systemic inflammation and antiretroviral (ARV) toxicity as well as viral infiltration of cardiac myocytes and intramyocardial lipids leading to myocardial fibrosis.^4,7–9

Although CVD is a leading cause of death in adults with HIV, the implications for disability are underrecognized. The primary determinant of exercise intolerance in adults with heart failure is diastolic dysfunction, which is characterized by elevated left ventricular (LV) filling pressure and impaired LV relaxation.¹⁰ Diastolic dysfunction predicts reduced cardiorespiratory fitness (CRF) in HFpEF¹¹ as well as heart failure with reduced ejection fraction (HFrEF).^12,13 Hypertension and metabolic disease contribute to the development of diastolic dysfunction via impaired LV relaxation that is associated with low CRF before diagnosis of HFpEF.^14,15

Adults with HIV have reduced CRF regardless of age and length of HIV infection.^16–18 Among adults with HIV, we previously reported that VO₂peak was 16% lower among those with hypertension.¹⁹ However, diastolic dysfunction is yet to be investigated as a mediator of low CRF in adults with HIV. The objective of this cross-sectional study was to examine the relationship of diastolic dysfunction with CRF in adults with HIV but no known heart disease or failure. We sought to determine the cardiac structural and functional determinants of low CRF.

Methods

Participants

Adults living with HIV who were receiving ARV were recruited from infectious disease clinics located at the Salem and Baltimore VA Medical Centers and Virginia Tech Carilion Clinic. Exclusion criteria included safety criteria for treadmill testing established by the American College of Sports Medicine (ACSM)²⁰ as well as recent AIDS-defining illness in the prior 6 months. Participants with history of atherosclerosis (CAD, myocardial infarction, stroke, and peripheral vascular disease), and diagnosis of New York Heart Association Classification III or IV heart failure, and severe lung disease (oxygen or systemic glucocorticoid steroid use) were also excluded. Written informed consent was obtained from participants and approved by each local institutional review board and the VA Research and Development Committee.

Cardiopulmonary exercise testing

Cardiopulmonary exercise testing (CPET) was performed on a treadmill using the modified Bruce protocol to evaluate CRF.²⁰ The test was terminated according to ACSM safety criteria or when the subject reported exhaustion. Open-circuit spirometry was used to measure gas exchange (oxygen and carbon dioxide) during treadmill testing. Oxygen utilization, carbon dioxide production, and minute ventilation values were collected breath-by-breath and averaged at 10-sec intervals with either a Vmax 29C (Sensor Medics) or True One 2400 (Parvo Medic) metabolic cart. CRF was defined as the peak oxygen utilization (VO₂peak), calculated as the average of the highest three 10-sec interval measurements near test termination. A cardiologist who was blinded to clinical data reviewed electrocardiogram tracings from the treadmill test at each site. Participants with evidence of cardiac ischemia were excluded from the study and referred for clinical evaluation.

Echocardiographic measures

Resting echocardiography, including two-dimensional and tissue Doppler imaging, was performed by an experienced cardiac sonographer who was blinded to participant data. The LV end diastolic and end systolic volumes were measured by the biplane modified Simpsons rule from apical 2- and 4-chamber views after manual tracing of the endocardial borders by a single experienced reader blinded to study data.²¹ Ejection fraction was calculated as the difference between end diastolic and end systolic volume divided by end diastolic volume. Pulsed wave Doppler examination of mitral inflow was performed and measured early (E) and late (A) mitral inflow velocities. The E/A ratio was calculated to assess LV relaxation. Deceleration time of early LV filling and isovolumetric relaxation time were also measured. Doppler tissue imaging of mitral annular motion was performed, which included the measurement of early diastolic annular velocity (e’) at both the medial (e'_m) and lateral (e'_L) positions. The medial measurement was used in the analysis and is represented hereafter as e’.²² Impaired myocardial relaxation was defined as e’ <8 cm/s.²³ The ratio of E/e’ was calculated as a measurement of LV filling pressure. Increased LV filling pressure was defined as E/e’ ≥10. LV mass was determined from anatomically correctly aligned linear measurements of LV cavity dimension and wall thickness. Diastolic function was classified as either normal or mild, moderate or severe impairment, based on previously defined criteria for E/A, E/e’, and deceleration time.²⁴ Participants with missing Doppler measurements or those who did not meet any of the defined criteria were designated as having diastolic function “not classified.” Pulmonary artery pressure (PAP) was estimated from the tricuspid regurgitation jet, which was visible in 53% (63/119) of echocardiograms. Pulmonary hypertension was defined as PAP ≥25 mmHG.²⁵

Systolic function was classified based on LV ejection fraction as normal (≥55%), mildly decreased (45–54%), moderately decreased (35–44%), and severely decreased (<35%).⁸ LV mass was calculated by biplane area/length method and indexed to body surface area. LV hypertrophy was defined as having LV mass index ≥116 g/m². Left atrial volume was calculated by the modified Simpson's rule and indexed to body surface area.²¹

Clinical characteristics

Information on comorbid conditions and medication was based on chart review and confirmed by history and physical examination at study enrollment. Values for CD4 lymphocyte count, plasma HIV-1 RNA level, and hemoglobin were extracted from the medical record or drawn at enrollment if not available in the prior 6 months. History of cigarette smoking, alcohol, and illicit drug use was obtained by participant self-report. Height and weight were measured to calculate body mass index (BMI, kg/m²). Anemia was defined as hemoglobin <12.0 g/dL.

Statistical analysis

The primary outcome was CRF measured as VO₂peak mL/(kg·min). Analysis of variance was used to determine the association of VO₂peak with demographic, clinical, and cardiac characteristics. Echocardiography measures were correlated with VO₂peak using Pearson or Spearman correlation coefficients. We performed a post hoc analysis on E/e’ that was stratified by age group based on the findings by Grewal et al.²⁶ For measures of diastolic function and clinical factors, which were associated with VO₂peak, the relationship of each measure with VO₂peak was then tested in a series of age-adjusted linear regression models in which each measure was included in a model that contained age. The independent contribution of each echo parameter to VO₂peak was determined in a multivariable model in which each echo parameter was included in a model that contained age and relevant clinical characteristics. Analyses were performed using Stata software (v14.0; StataCorp, College Station, TX). All analyses were two tailed.

Results

A total of 136 individuals living with HIV provided informed consent for the study. Participants were excluded if they were not receiving ARV (n = 9) or failed to complete both CPET and echocardiogram (n = 8). The resulting 119 participants represent the analytical set of this study. Table 1 shows the demographic and clinical characteristics. The majority were African American (86%) with a mean [standard deviation (SD)] age of 56.6 (7.1) years. The median [interquartile range (IQR)] BMI was 25.2 (22.9–29.5) kg/m². The median (IQR) CD4 lymphocyte count was 492 (300–729) cells/mm³. Mean (SD) hemoglobin was 13.7 (1.3) g/dL.

Table 1.

Characteristics of Study Population and Test for Association with Mean VO₂peak

Characteristics	Participants n = 119 n (%)	VO2peak, mL/(kg·min) mean (SE)
Age, years
≥50	106 (89.0)	25.3 (0.4)*
<50	13 (10.9)	32.2 (1.8)
Race
Black	103 (86.5)	26.9 (1.7)
Other	16 (13.4)	25.9 (0.5)
Gender
Male	115 (96.7)	26.2 (0.5)
Female	4 (3.3)	22 (2.2)
BMI, kg/m2
Overweight and obese	65 (54.6)	25.4 (0.7)
Normal and underweight	54 (45.3)	26.8 (0.6)
Hypertension
Present	88 (73.9)	25.4 (0.5)†
Not present	31 (26.1)	27.8 (1.0)
Diabetes
Present	30 (25.2)	22.9 (0.8)*
Not present	89 (74.8)	27.1 (0.5)
Anemia (hemoglobin <12.0 g/dL)
Present	33 (27.7)	24.5 (1.0)
Not present	86 (72.3)	26.6 (0.5)
Smoking status
Never smoked	21 (17.6)	27.6 (1.2)
Former smoker	31 (26.8)	24.7 (1.0)
Current smoker	66 (55.4)	26.1 (0.6)
Injection drug use
Never	28 (23.7)	27.0 (1.0)
Prior	66 (55.9)	25.1 (0.6)
Current	24 (20.3)	27.4 (1.2)
CD4 lymphocyte count
<500 cells/mm3	61 (51.3)	25.7 (0.7)
≥500 cells/mm3	58 (48.7)	26.4 (0.6)
HIV-1 RNA
>75 copies/mL	16 (13.5)	26.1 (1.5)
≤75 copies/mL	103 (86.5)	26.0 (0.5)
History of AIDS defining condition
Present	56 (47.1)	26.8 (0.8)
Not present	63 (52.9)	25.3 (0.6)
Current ARV regimen:
Nucleoside reverse transcriptase inhibitor
Yes	86 (73.5)	26.0 (0.6)
No	31 (26.5)	26.2 (0.9)
Non-nucleoside reverse transcriptase inhibitor
Yes	18 (15.1)	23.3 (1.0)†
No	101 (84.9)	26.5 (0.5)
Protease inhibitor
Yes	62 (52.1)	26.7 (0.7)
No	57 (47.9)	25.2 (0.6)
Integrase inhibitor
Yes	15 (12.6)	24.6 (1.3)
No	104 (87.3)	26.2 (0.5)
LV diastolic function
Normal	64 (53.7)	27.2 (0.6)†
Abnormal (mild/moderate/severe)	45 (37.8)	24.2 (0.7)
Unclassified	10 (8.4)	27.2 (2.0)
e’, cm/s
Normal	41 (36.2)	27.7 (0.9)†
Abnormal (<8 cm/s)	72 (63.7)	25.2 (0.6)
E/e’
Normal	82 (68.9)	26.6 (0.6)
Abnormal (≥10)	37 (31.1)	24.8 (0.8)
Pulmonary hypertension‡
Present (PAP ≥25 mmHg)	27 (42.8)	25.4 (0.9)
Not present	36 (57.1)	27.0 (1.0)
LV mass index, g/m2
Normal	107 (89.9)	26.1 (0.5)
LVH (≥116 g/m2)	12 (10.1)	25.6 (1.2)
Ejection fraction, %
Normal	108 (90.8)	26.2 (0.5)
Abnormal	11 (9.2)	23.8 (1.3)

Differences between groups by analysis of variance.

^*p-value <0.01.

^†p-value <0.05.

^‡n = 63.

ARV, antiretroviral; BMI, body mass index; LVH, left ventricular hypertrophy; PAP, pulmonary artery pressure.

CPET showed a mean (SD) VO₂peak of 26.1 (5.5) mL/(kg·min) and 2.11 (0.55) L/min. The mean (SD) duration of the test was 12.1 (2.3) min. Respiratory exchange ratio (RER) median (IQR) was 1.11 (1.03–1.16). The majority (86%) of participants reached RER >1.05 or 85% of age-predicted maximum heart rate. Seven participants were receiving a beta-blocker medication. VO₂peak was inversely correlated with age (r = −0.40, p < 0.01) and positively correlated with hemoglobin concentration (r = 0.24, p < 0.01). Table 1 shows the relationship of participant characteristics with VO₂peak [mL/(kg·min)]. VO₂peak was significantly lower in patients with hypertension and diabetes and trended lower in those with anemia (p = 0.06). After adjustment for age, diabetes and anemia were significantly associated with VO₂peak, but hypertension was not. CD4 lymphocyte count, plasma HIV-1 level, and ARV drug class were not associated with VO₂peak.

Diastolic function was graded as normal (n = 64, 53.7%), mild dysfunction (n = 17, 14.3%), and moderate dysfunction (n = 28, 23.5%). Severe dysfunction was not found. Diastolic function was not classified in 10 (8.4%) individuals, due to missing data for one echo parameter or failure to fit the predetermined definitions for E/A, deceleration time, and E/e’.²⁴ VO₂peak was significantly lower if any diastolic dysfunction was present (Table 1). Impaired myocardial relaxation (e’ <8 cm/s) was present in 64% of the participants and was significantly associated with a lower mean VO₂peak compared with those with normal myocardial relaxation. There was a trend for lower mean VO₂peak among those with elevated LV filling pressure (E/e’ ≥10; p = 0.1). Intraventricular septal width inversely correlated with VO₂peak. However, LV mass and LV mass index were not significantly associated with VO₂peak.

Systolic function was classified as normal in most of the participants (91%) and the median (IQR) ejection fraction was 60.5% (56.1–64.5). There was a trend for lower mean VO₂peak in the 11 individuals with ejection fraction <55% (Table 1, p = 0.1). PAP did not correlate with VO₂peak among the 63 with an estimated PAP (r = −0.04, p = 0.7). Pulmonary hypertension was present in 43% (27/63) and was not associated with VO₂peak (Table 1).

Clinical factors associated with diastolic dysfunction were age and diabetes. Compared with those without diastolic dysfunction, participants with diastolic dysfunction were older (58.8 [1.1] vs. 55.3 [0.9] years, p = 0.01) and had a higher prevalence of diabetes (35.6% vs. 18.8%, p = 0.045). Surprisingly, the prevalence of hypertension was similar (73.3% vs. 75.5%, p = 0.9). HIV-related characteristics, including ARV drug class, were not associated with diastolic dysfunction. The prevalence of diastolic dysfunction was not significantly different between participants with and without a protease inhibitor-based ARV (32% vs. 44%, p = 0.2).

The correlation of cardiac characteristics with VO₂peak is provided in Table 2. LV relaxation (E/A) and myocardial relaxation (e’) positively correlated with VO₂peak. The inverse correlation of LV filling pressure (E/e’) with VO₂peak was not significant (p = 0.1). However, age-stratified analysis using a cutoff of 60 years showed a significant association of E/e’ with VO₂peak in younger (n = 80) participants (r = −0.23, p = 0.04) compared with older (n = 32) participants (r = 0.09, p = 0.6). Further, elevated LV filling pressure (E/e’ ≥10) was associated with significantly lower VO₂peak compared with those with normal pressure [Δ −3.4 mL/(kg·min), p = 0.02] in younger participants.

Table 2.

Correlation of Cardiac Function and Structure with VO₂peak [mL/(kg·min)]

Cardiac characteristics	Median (range)	Correlation coefficient with VO2peak
Diastolic function
E/A	0.9 (0.5–2.5)	0.28*
e’, cm/s	7.2 (3.2–15.0)	0.23†
E/e’	8.5 (2.6–15.6)	−0.15
Deceleration time, ms	216.0 (143.0–495.0)	−0.10
Isovolumic relaxation time, ms	88.0 (48.0–137.0)	−0.07
Systolic function
Ejection fraction, %	60.6 (26.12–73.0)	<0.01
Cardiac structure
LV mass, g	169.0 (81.9–279.5)	−0.17
LV mass index, g/m2	84.9 (48.6–145.5)	−0.05
Intraventricular septal width, cm	1.0 (0.7–1.5)	−0.25*
Left atrial area, cm2	17.6 (10.5–42.5)	0.18†
Left atrial volume index, mL/m2	30.6 (9.7–65.7)	0.26†

Pearson or Spearman correlation.

^*p-value <0.01.

^†p-value <0.05.

In linear regression models, myocardial relaxation and LV relaxation were associated with VO₂peak independent of age (Table 3). Impaired myocardial relaxation (e’ <8 cm/s) was associated with an average 1.9 mL/(kg·min) lower VO₂peak, adjusted for age and clinical factors. Each unit increase in LV relaxation (E/A) was associated with an average 4.4 mL/(kg·min) higher VO₂peak, adjusted for age and clinical factors. LV filling pressure (E/e’) was not significantly associated with VO₂peak with adjustment for age and clinical factors. Diastolic dysfunction explained 28–31% of the variance in VO₂peak.

Table 3.

Linear Regression of VO₂peak [mL/(kg·min)] and Measures of Left Ventricular Relaxation (E/A), Impaired Myocardial Relaxation (e’ <8 cm/s), and Increased Left Ventricular Filling Pressure (E/e’ ≥10)

Independent variable	Age-adjusted model		Multivariable model*
	β-coefficient	Adjusted R2	β-coefficient	Adjusted R2
E/A	4.4†	0.20	4.4†	0.31
e’ <8, cm/s	−2.1‡	0.19	−1.9‡	0.29
E/e’ ≥10	−1.5	0.17	−1.1	0.27

^*Includes age, anemia, diabetes, and hypertension.

^†p-value <0.01.

^‡p-value <0.05.

Discussion

We found that diastolic dysfunction is associated with low CRF in adults living with HIV, independent of age and clinical factors. Our results are consistent with the general literature but provide novel findings in adults with HIV. Given that an increase in VO₂peak of 1 metabolic equivalent [MET, 3.5 mL/(kg·min)] is associated with a 13% reduced risk in all-cause mortality and 15% reduced risk in CVD events,²⁷ our findings have important clinical implications.

HFpEF constitutes approximately half of the diagnoses of congestive heart failure in older adults^24,28 and carries the same mortality risk as HFrEF.²⁹ In a longitudinal cohort of Veterans without baseline CVD, adults with HIV had a 21% increased risk of HFpEF.⁵ In our similar sample of adults with HIV, we found that about a third of participants had either mild or moderate diastolic dysfunction. This prevalence is consistent with other cross-sectional studies of middle-aged adults with HIV.^2,8,30 Severe dysfunction, reported in <5% of adults with HIV,³¹ was absent in our participants and likely reflects the exclusion of those with heart failure. Our findings establish the association of diastolic dysfunction with CRF in adults with HIV for the first time to our knowledge. In participants with diastolic dysfunction, the mean VO₂peak was 3.0 mL/(kg·min) (0.85 METs) lower compared with those with normal function.

The primary manifestation of diastolic dysfunction in our participants was impaired myocardial relaxation, present in 64% of our sample. In comparison with the larger studies of patients with HFpEF,³² the correlation between myocardial relaxation (e’) and VO₂peak was modest in our study. However, using an established threshold for impaired myocardial relaxation (e’ <8 cm/s), we found a clinically significant reduction in VO₂peak. Likewise, preserved LV relaxation (E/A) has a positive effect on VO₂peak. For each unit increase in E/A, VO₂peak was on average 4.4 mL/(kg·min) higher (1.3 METs), independent of age and clinical characteristics. In contrast, LV filling pressure (E/e’) was not significantly associated with VO₂peak in our participants unless we stratified by age. Among the younger participants, elevated LV filling pressure (E/e’ ≥10) was associated with a lower VO₂peak of 1 MET compared with those with normal pressure. There was no association in the 32 older participants. In a large observational study of almost 3000 patients referred for exercise echocardiography and not limited by ischemia, Grewal et al. showed that increased LV filling pressure (E/e’) was associated with reduced fitness.²⁶ Interestingly, they reported a significant interaction between age and LV filling pressure. Among those with elevated LV filling pressure, the decrease in VO₂peak with increased age was greater. In this context, our findings suggest an interaction between age and HIV in the relationship between LV filling pressure and fitness, which needs to be investigated further.

Among adults living with HIV, Thoni et al. showed that diastolic dysfunction was associated with impaired cardiac response to exercise (cardiac output) in 16 men with HIV but did not test the relationship with CRF.³³ A larger study of diastolic function, which included adults with and without HIV, used treadmill testing to screen for ischemia, but did not test the association of diastolic dysfunction with CRF.³ Cardiac research in adults with HIV shows that risk factors for diastolic dysfunction are similar to adults without HIV and include metabolic disease,³⁰ obesity,⁶ and hypertension.^4,34 Each of these conditions is related to low CRF, yet the role of diastolic dysfunction has not been previously reported in adults with HIV. Our findings demonstrate that the relationship between diastolic function and CRF is independent of the effects of diabetes and hypertension.

Myocardial fibrosis and stiffening are important mechanisms of diastolic dysfunction, although peripheral determinants also play a role.^11,35 In adults with HIV, cardiac MRI shows that myocardial fibrosis is associated with intramyocardial lipid content, which positively correlates with the duration of ARV exposure but not specific subclasses of ARVs.^9,36 However, preclinical research supports several mechanisms by which specific subclasses of ARV could induce cardiac fibrosis.³⁷ Protease inhibitors, also linked with increased risk of HFpEF,⁵ may affect platelet activation and profibrotic signaling pathways.³⁷ HFpEF can be considered a geriatric syndrome driven by age-related chronic inflammation.³⁸ The ongoing Characterizing Heart Function on Antiretroviral Therapy (CHART) study will help differentiate the role of HIV infection, ARV, and systemic inflammation in cardiac fibrosis and the development of diastolic dysfunction.³⁹ A current pitavastatin trial in adults with HIV will determine the impact on CVD and inflammatory biomarkers⁴⁰ and may provide insight in its application for diastolic dysfunction. However, statins are not recommended for HFpEF in general⁴¹ and are underprescribed in adults with HIV.⁴² There are few pharmacologic options for treatment of HFpEF and current recommendation focuses on underlying comorbid conditions, including diabetes.⁴¹ Diabetes is an independent risk factor for HFpEF⁴³ driven by inflammation-mediated cardiac fibrosis.¹⁵ Our findings in adults with HIV show that diabetes is a risk factor for diastolic dysfunction, which is associated with a reduced VO₂peak of 1 MET. Given the paucity of pharmacologic interventions for diastolic dysfunction, this raises the question of the role of exercise training. The positive effect of exercise on both metabolic disease and inflammation is well-established. However, improvement in diastolic function per se has been less consistent but observed in a few exercise trials.^44,45 In the Exercise training in Diastolic Heart Failure (EX-DHF) trial, increased VO₂peak in the exercise group was associated with decreased LV filling pressure and decreased levels of procollagen type I levels.⁴⁶

The primary study limitation is that our sample consisted predominantly of African American men who had no history of CAD or heart failure. Generalization to other patient groups is therefore limited. However, we expect that the tight eligibility criteria would bias our results to the null, and future studies of HIV and HFpEF will have even more robust findings. The cross-sectional nature of this study prohibits conclusions regarding the causal relationship between diastolic dysfunction and low CRF and the mediating role of age with comorbidity. Finally, while we used established echocardiography criteria for diastolic dysfunction,²⁴ these measures are not precise and other mechanisms besides diastolic dysfunction can affect them, such as pulmonary hypertension. For this reason, we did not focus on the significant findings with the left atrial volume index.⁴⁷ While our findings did not show an association between pulmonary hypertension and VO₂peak, echocardiographic estimate of PAP was available in only 50% of the study population, which excluded those with severe lung disease. Further research in the central determinants of impaired fitness, which include pulmonary function, is needed to differentiate the contribution of cardiac and pulmonary limitations.

In conclusion, our findings demonstrate the relationship between diastolic dysfunction and VO₂peak in adults with HIV. While many middle-aged adults with diastolic dysfunction are asymptomatic, our results emphasize the physiologic implications of diastolic dysfunction. In the next decade, the number of adults with HIV and HFpEF will increase. Ideally, biomarkers will be able to identify at-risk individuals before the development of cardiac steatosis and fibrosis. In the meantime, clinical strategies need to include prevention and treatment of modifiable factors that impact both diastolic function and CRF.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This research was supported by The Department of Veterans Affairs, Veterans Health Administration, Rehabilitation Research and Development Service (I01 RX000667) and VA Senior Research Career Scientist Award (ASR)), the NIA Claude D. Pepper Older Americans Independence Center (P30AG028747), and R01HL095136.