Stages of Valvular Heart Disease Among Older Adults in the Community: The Atherosclerosis Risk in Communities Study

Abstract

Background

Limited data exist regarding American College of Cardiology/American Heart Association (ACC/AHA) valvular heart disease (VHD) stage prevalence, progression, and association with incident cardiovascular diseases in late life.

Methods

Participants in the Atherosclerosis Risk in Communities (ARIC) prospective community-based cohort study underwent protocol echocardiography at ARIC Visits 5 (2011–2013) and 7 (2018–2019) and their aortic stenosis, aortic regurgitation, mitral stenosis, and mitral regurgitation stage was defined based on ACC/AHA guidelines. The overall VHD stages prevalence at Visit 5 was measured. The association between VHD stages and incident adjudicated death, heart failure (HF), coronary heart disease (CHD), stroke, and atrial fibrillation (AF); using Cox proportional hazard models adjusted for age, sex, race, hypertension, diabetes, prior myocardial infarction, HF, body mass index, study center, systolic blood pressure, estimated glomerular filtration rate, and low-density lipoprotein at Visit 5. longitudinal changes in VHD stages prevalence over approximately 6 years were estimated using inverse probability of attrition weights to account participant attrition.

Results

Among 6,118 ARIC participants, mean±SD age was 76±5 years, 42% were male, and 22% reported Black race. Stage A VHD was present in 39%, Stage B in 17%, and Stage C/D in 1.1%, while 0.7% had previously undergone valve replacement or repair. A graded association was observed between Stage A, B, and C/D VHD and risk of all-cause mortality, incident HF, incident AF, and incident CHD, but not incident stroke. Similar findings were observed for stages of each valvular lesion individually. During the 6.6 [IQR, 6.1–7.0] years between Visit 5 and Visit 7 (mean age 81±4 years), the prevalence of freedom from VHD stage decreased from 43% to 24%, while the prevalence of Stage C/D VHD increased from 1% to 7% respectively.

Conclusions and Relevance

Subclinical VHD is common in older adults, with 39% at risk (Stage A) and 17% with progressive VHD (Stage B), and is independently associated with risk of incident cardiovascular events. VHD stages progress over six years in late-life, with a several-fold increase in prevalence of severe VHD (Stage C/D), highlighting the public health importance of interventions to mitigate VHD progression.

Introduction

Valvular heart disease (VHD) is associated with significant morbidity and mortality and demonstrates a marked increase in prevalence with advancing age. While the prevalence of moderate to severe VHD is estimated at 2.5% in the general population, this estimate increases to 13.2% among those >75 years old.¹ Lesser degrees of valvular disease are even more common, occurring in approximately 51% of community-dwelling persons ≥65 years of age.² The burden of VHD is expected to grow substantially as the population ages, with persons >65 years old anticipated to account for 20% of the US population by 2030.³

VHD is progressive, beginning with structural alterations in valve morphology, moving through increasing degrees of valvular dysfunction (stenosis or regurgitation), and ultimately culminating in severe symptomatic disease. In 2014, the American College of Cardiology /American Heart Association (ACC/AHA) adopted the conceptual framework of VHD stages to emphasize its progressive nature.^4,5 This schema defines VHD stages as: Stage A – at risk for valve dysfunction; Stage B – progressive valvular dysfunction; Stage C – severe asymptomatic valve dysfunction; and Stage D – severe symptomatic valve dysfunction. However, to our knowledge, no community-based estimates exist for the prevalence of VHD stages, or their progression over time, particularly in late-life when the burden of VHD is greatest. We aimed to (1) describe the prevalence of VHD stages among older adults in the community; (2) determine the prognostic relevance of VHD stages for incident cardiovascular diseases (CVD); and (3) characterize the progression in VHD stages over approximately 6 years in late life.

Methods

The ARIC study has been approved by institutional review boards of all participating institutions, and all participants provided written informed consent. Data availability and detailed policies for requesting ARIC data can be found at https://www2.cscc.unc.edu/aric/pubs-policies-and-forms-pg. ARIC data can also be obtained from the NHLBI BioLINCC repository (https://biolincc.nhlbi.nih.gov/home/).

Study Population

The Atherosclerosis Risk in Communities (ARIC) study is a prospective epidemiologic cohort study, the design and methods of which have been previously described.⁶ Between 1987 and 1989, 15,792 middle-aged subjects were enrolled in 4 communities in the United States: Forsyth County, NC, Jackson, MS, suburban Minneapolis, MN, and Washington County, MD. Of the 10,742 alive at the time of Visit 5 (2011 to 2013), 6,118 participants attended and underwent echocardiography with adequate images for assessment of VHD. Of these, 4,895 were alive at the time of Visit 7 (2018 to 2019), 2,896 of whom attended Visit 7 and underwent protocol echocardiography (Figure 1).

Figure 1.

Flowchart of study participants. The bold text box represents the study population, left side boxes show the mortality in the corresponding right side boxes. VHD indicates valvular heart disease.

Echocardiography and Definition of Valvular Heart Disease Stages

Procedures for echocardiography in ARIC at Visit 5, including reproducibility metrics, have been previously described, and were similar at Visit 7.⁷ At both Visits 5 and 7, studies were acquired by certified sonographers using uniform imaging machines (Philips iE33, Koninklijke Philips, The Netherlands) and probes (Philips XMatrix) and acquisition protocols. Quantitative measures for studies from both visits were performed at the same dedicated Echocardiography Reading Center by trained analysts who were blinded to clinical information and in accordance with American Society of Echocardiography (ASE) recommendations.^8–10 At both Visits, all quantitative measures were over-read by study investigators who were staff cardiologists at the Brigham and Women’s Hospital with Core Cardiology Training Symposium (COCATS) level III advanced training in echocardiography and/or ASE Board Certification in Comprehensive Adult Echocardiography.

VHD stages were defined uniformly at both Visits 5 and 7 based on the ACC/AHA guideline recommendations and operationalized in this study as shown in Table 1 (see Data Supplement for additional details). Concordant with ACC/AHA VHD guidelines, mitral regurgitation (MR) was quantified based on the MR jet area (MRJA) to left atrial area (LAA) ratio, aortic stenosis (AS) was quantified based on aortic valve (AV) peak jet velocity (Vmax) and AV area (AVA), aortic regurgitation (AR) was qualitatively assessed by over-reading echocardiographers based on color Doppler signal, and mitral stenosis (MS) was assessed based on mitral valve area (MVA) calculated from the pressure half-time in addition to qualitatively assessment by over-reading echocardiographers (see Data Supplement for additional details). Among participants who attended Visit 5 but not Visit 7, those with a post-Visit 5 hospitalization or death certificate with a VHD International Classification of Diseases (ICD) code (see Data Supplement for specific codes) were classified as VHD Stage C/D at Visit 7. Those with an ICD procedure code for a VHD intervention post-Visit 5 were categorized as having a valve replacement or repair at Visit 7.

Table 1:

Operationalization of VHD stages in ARIC

	Aortic stenosis	Aortic regurgitation	Mitral stenosis	Mitral regurgitation
Stage A: At risk	Sclerosis: Vmax from 1.5 to <2.0m/sec     or Bicuspid valve	Sclerosis: Vmax from 1.5 to <2.0m/sec     or Bicuspid valve	Mitral annular calcification     or Rheumatic features: rare	Mitral annular calcification     or Mitral valve prolapse (MVP)     or Mild MR: (^MRJA:LAA ratio 5 –20%)
Stage B: Progressive	Mild AS: Vmax from 2.0 to <3 m/sec     or Moderate AS: Vmax from 3 to <4 m/sec or mean ΔP from 30 to< 40 mmHg	Mild AR: (Qualitative)     or Moderate AR: (Qualitative)	Qualitative MS and $Calculated MVA >1.5 cm2	Moderate MR: (^MRJA:LAA ratio 20–40%) or eccentric jet with mild M.R.)
Stage C: Asymptomatic Severe	Severe AS:(Vmax ≥4.0 m/sec or mean ΔP ≥ 40 mmHg) + Stage C1: LVEF ≥50%     or Stage C2: LVEF <50%     or Asymptomatic low flow*	Severe AR (Qualitative) + Stage C1: LVEF ≥55%, and LVESD ≤5 cm     or Stage C2: LVEF <55% or LVESD > 5 cm	Qualitative MS and $Calculated MVA ≤ 1.5 cm2	Severe MR: (^MRJA:LAA ratio ≥40% or eccentric jet with moderate M.R.) + Stage C1: LVEF>60% and LVESD <4.0cm     or Stage C2: LVEF ≤60% or LVESD≥4.0cm
Stage D: Symptomatic Severe	Symptoms* or Angina + Stage D1: Severe AS ± LVEF ≥50%     or Stage D2: low flow and LVEF <50%,     or Stage D3: low flow and LVEF ≥50%	Symptoms* + Severe AR (Qualitative)	Symptoms* + $Calculated MVA ≤ 1.5 cm2	Symptoms* + Severe MR

V_max: Peak aortic valve velocity by Doppler echocardiography, Mean ΔP: The mean pressure gradient across the aortic valve by Doppler.

^*Low flow: V_max < 4.0 m/sec and AVA ≤1.0cm².

^*Symptoms: dyspnea, exhaustion, or heart failure

^$Calculated Mitral valve area (MVA) = 220 / Pressure half-time (PHT) where Pressure half-time (PHT) = mitral inflow deceleration time × 0.29

^{^}MRJA:LAA ratio: the ratio between mitral regurgitation jet area and left atrial area

Prevalent and Incident Cardiovascular Events

ARIC cohort participants undergo surveillance for cardiovascular events through annual questionnaires and review of hospitalization discharge codes as previously described.⁶ Coronary heart disease (CHD), heart failure (HF), and stroke were ascertained based on medical record abstraction and committee adjudication of hospitalizations while atrial fibrillation (AF) events were based on hospitalization ICD discharge codes (see Data Supplement for additional details).^6,11–17 Death was ascertained using the National Death Index. For incident CHD, HF, stroke, and death, the end date for follow-up was December 31, 2019, except for 697 participants from Jackson center whose follow-up was through December 31, 2017 due to administrative reasons. For incident AF the end date for follow-up was December 31, 2017 for all study participants.

Clinical Covariates and Biomarkers

Hypertension was ascertained based on participant report of blood pressure medication use or blood pressure ≥140/90 mmHg at any ARIC visit. Diabetes mellitus was ascertained based on self-report of a physician diagnosis of diabetes mellitus, antidiabetic medication use, fasting glucose ≥126 mg/dL, or non-fasting glucose ≥200 mg/dL at any ARIC visit. Body mass index was calculated from weight and height assessed at Visit 5. Frailty was assessed at Visit 5 using Fried criteria, which incorporates gait speed, grip strength, low energy expenditure, weight loss, and exhaustion.¹⁸ Estimated glomerular filtration rate estimated (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation. ¹⁹ N-terminal pro-brain natriuretic peptide (NT-proBNP) was measured using electrochemiluminescent immunoassay (Roche Diagnostics), with a lower detection limit of ≤5 ng/mL. ²⁰ High sensitive Troponin T (hs-TnT) was measured using a highly sensitive assay (Elecsys Troponin T; Roche Diagnostics, Indianapolis, IN), and the limit of the blank was 3 ng/L.^20,21

Statistical Methods

Participants were classified based on the most severe valve stage of the following valve lesions: AS, AR, MS, MR. VHD stage prevalence was also described stratified by age category (65 – 70, 71 – 75, 76 – 80, >80 years old), sex, and race. Multivariable linear regression models were used to relate VHD stages to concentrations of NT-proBNP and hs-TnT at Visit 5. Values of both biomarkers were log-transformed to achieve normality. Multivariable Cox proportional hazard models were used to assess the relationship of the VHD stage at Visit 5 with incident CHD, HF, AF, stroke, or death. Initial models adjusted for age, sex, and race. Subsequent models further adjusted for hypertension, diabetes, prior myocardial infarction (MI), HF, body mass index, Field Center, systolic blood pressure, eGFR and low-density lipoprotein (LDL) at Visit 5. For each endpoint, models excluded participants with the prevalent condition at Visit 5, and participants with prior valve replacement. Analyses were performed by VHD stage overall and for stage of each valvular lesion individually (AS, AR, MS, and MS). In a sensitivity analysis, we performed a competing risk analysis to account for the competing risk of death for the nonfatal endpoints. To assess the potential impact of non-random Visit 5 non-attendance on the survival analyses, we performed a sensitivity analysis incorporating inverse probability of attrition weights (IPAW; see Data Supplement for additional details).^22,23

To assess transitions in VHD stage between Visit 5 and Visit 7 overall, and by the valvular lesion, we employed IPAW to account for Visit 7 non-attendance among ARIC Visit 5 participants. Visit 7 non-attendance was modeled among participants alive through the end of Visit 7 using the following covariates from Visit 5: age, gender, race, study center, frailty, cancer, diabetes, hypertension, and prevalent HF. The resulting calculated weights were incorporated to estimate VHD stage prevalence at Visit 7 among all Visit 5 participants alive at the time of Visit 7.

All analyses were performed using STATA 16. Two-sided P-values of less than 0.05 were considered significant.

Results

Prevalence and Correlates of VHD Stages at Study Visit 5

The mean age of the study sample at Visit 5 was 76±5 years, 42% were male, and 22% reported Black race (Table 2). Stage A VHD was present in 38.6%, Stage B in 16.5%, and Stage C/D in 1.1%, while 0.7% had previously undergone valve replacement or repair (Figure 2A). VHD was absent in 43.2 % of participants. Older age was associated with a higher prevalence of all VHD stages (Figure 2B). The prevalence of VHD was 44.0% and 53.5% in Black men and women respectively, and was 56.3% and 60.6% in White men and women respectively (Figure 2C). Greater VHD stage was characterized by a higher prevalence of cardiovascular risk factors and of prevalent CVD (Table 2). In fully adjusted analyses, greater VHD stage at Visit 5 was associated with higher concentrations of NT-proBNP and hs-TnT, including both Stage A and in Stage B compared to participants free of VHD (Figure S1). Similar associations were observed within each valvular lesion (Figure S2).

Table 2:

Baseline characteristics at Visit 5 by VHD stage

	Overall	No VHD	Stage A	Stage B	Stage C/D
N	6118	n=2640	n=2362	n=1010	n=66	P-value
Age, years	76 ± 5	75 ± 5	76 ± 5	77 ± 5	78 ± 5	<0.001
Male, n (%)	2576 (42%)	1168 (44%)	910 (39%)	440 (44%)	32 (48%)	0.23
Black, n (%)	1334 (22%)	684 (26 %)	483 (20%)	148 (15%)	11 (17%)	<0.001
Center, n (%)						0.1
Forsyth County, NC	1416 (23%)	541 (20%)	586 (25%)	263 (26%)	21 (32%)
Jackson, MS	1214 (20%)	624 (24%)	439 (19%)	134 (13%)	9 (14%)
Minneapolis, MN	1823 (30%)	851 (32%)	628 (27%)	316 (31%)	13 (20 %)
Washington County, MD	1665 (27%)	624 (24%)	709 (30%)	297 (29%)	23 (35%)
Ever Smoker, n (%)	3764 (62%)	1666 (63%)	1388(59%)	648 (64%)	38 (58%)	0.49
Current Smoker, n (%)	349 (6%)	158 (6 %)	147 (6 %)	40 (4 %)	3 (5 %)	0.07
HTN, n (%)	5103 (83%)	2142 (81%)	2003 (85%)	864 (86 %)	58 (88%)	<0.001
DM, n (%)	2301 (38%)	1048 (40%)	884 (37%)	322 (32%)	32 (48%)	<0.001
CKD, n (%)	1705 (28%)	659 (25%)	677 (29%)	329 (33%)	24 (37%)	<0.001
CHD, n (%)	939 (16%)	314 (12%)	395 (17%)	193 (19%)	18 (28%)	<0.001
MI, n (%)	709 (12%)	261 (11%)	285 (13%)	138 (14%)	14 (23%)	<0.001
HF, n (%)	955 (16%)	328 (12%)	389 (16%)	192 (19%)	18 (27%)	<0.001
Stroke, n (%)	224 (4%)	69 (3 %)	98 (4%)	49 (5 %)	3 (5 %)	<0.001
AF, n (%)	442 (7%)	120 (5%)	191 (8 %)	103 (10 %)	10 (15 %)	<0.001
BMI, kg/m2	29 ± 6	29 ± 6	29 ± 6	28 ± 5	28 ± 5	<0.001
SBP, mmHg	130 ± 18	129 ± 17	131 ± 18	132 ± 19	132 ± 20	<0.001
DBP, mmHg	67 ± 11	67 ± 10	66 ± 11	65 ± 11	65 ± 15	<0.001
Pulse Pressure, mmHg	64 ± 15	61 ± 13	65 ± 15	67 ± 16	67 ± 15	<0.001
Heart rate, bpm	65 ± 11	66 ± 11	65 ± 11	64 ± 11	64 ± 11	<0.001
Hemoglobin A1c (%)	5.9 ± 0.8	6.0 ± 0.9	5.9 ± 0.8	5.8 ± 0.7	6.0 ± 0.9	<0.001
eGFR,mL·min-1−1.73m−2	70 ± 17	71 ± 17	70 ± 17	67 ± 18	65 ± 18	<0.001
LDL (mg/dL)	104 ± 35	106 ± 35	103 ± 33	103 ± 35	99 ± 34	0.007
HDL (mg/dL)	52 ± 14	52 ± 14	53 ± 14	52 ± 13	51 ± 15	0.88
NT-proBNP (ng/l)	134[68,266]	101[54,189]	149[77,300]	213[100,449]	297[141,823]	<0.001
Hs-Troponin T(ng/l)	11[7,16]	10[7,15]	11[7,16]	12[8,19]	15[9,25]	<0.001

Hypertension (HTN), diabetes mellites (DM), chronic kidney disease (CKD), coronary heart disease (CHD), myocardial infarction (MI), heart failure (HF), atrial fibrillation (AF), body mass index (BMI), systolic blood pressure (SBP), Diastolic blood pressure (DBP), estimated glomerular filtration rate (eGFR), low density lipoprotein (LDL), high density lipoprotein (HDL). Trends in baseline characteristics across groups were compared using linear regression, chi-squared trend tests, and Cuzick’s non-parametric trend test, as appropriate.

Figure 2.

Prevalence of VHD stages in ARIC at Visit 5 (n=6,118; mean age 76±5 years) overall (Panel A), by age category (Panel B), by race and gender groups (Panel C), and by valve lesion (Panel D). Prevalence by race and gender group is age-adjusted. MR indicates mitral regurgitation; MS, mitral stenosis; AR, aortic regurgitation; AS, aortic stenosis and VHD, valvular heart disease.

Stages of Aortic Valve Disease

Stage A AS was present in 15.4% (bicuspid valve in 11 participants), Stage B in 4.1% (3.8% mild AS, 0.3% moderate AS), Stage C (asymptomatic severe AS) in 27 participants (0.4%), and Stage D (symptomatic severe AS) in 26 participants (0.4%), while 32 participants had undergone prior AV replacement (Figure 2D). Stage C/D AS was responsible for 80% of participants with Stage C/D VHD overall. No participants had severe (Stage C) AR. Stage B AR was responsible for 62% of participants with Stage B VHD overall. Mild AR represented 92% of Stage B AR.

Stages of Mitral Valve Disease

Stage A MR was present in 38.6% (mild MR in 28.5%), Stage A MR was responsible, either alone or in combination with other Stage A lesions, for 80% of the participants who had overall VHD Stage A. MS was rare in this sample, with rheumatic deformity identified in only 1 participant, and qualitative mild stenosis related to calcification noted in 11 participants (0.2% prevalence; only 1 with calculated MVA ≤ 1.5 cm²).

VHD Stages and Incident Cardiovascular Events

Over a median follow-up of 6.5 (IQR 3.7–7.7) years, 1,295 participants died, incident HF occurred in 553 of those free of HF at Visit 5, incident CHD in 300 of those free of CHD at Visit 5, and incident stroke in 250 of those free of stroke at Visit 5. Over a median follow-up of 5.5 (IQR 4.8–5.9) years, incident AF occurred in 564 of those free of AF at Visit 5. In models adjusting for demographics and cardiovascular co-morbidities, a graded association was observed between Stage A, B, and C/D VHD and risk of all-cause mortality, incident HF, incident AF, and incident CHD, but not with risk of incident stroke (Figure 3, Table 3). Notably, compared to those free of VHD, each stage was associated with a heightened risk of these outcomes in adjusted models, including Stage A (Figure S3). Similar associations were observed after incorporating IPAW to account for Visit 5 non-attendance (Table S1), and in a sensitivity analysis accounting for the competing risk of death (Table S2). Similar associations were also observed for stages of each valvular lesion (Table S3, Figure S4), and after excluding participants who had another concomitant valvular lesion with a higher stage (Figure S5)

Figure 3.

VHD stages and incident events: Kaplan-Meier curves. The population at risk of each outcome is shown in a table below the curves at three-time intervals. *P-value of Log-rank test. CHD indicates coronary heart disease; AF, atrial fibrillation; HF, heart failure; and VHD, valvular heart disease.

Table 3:

Association of VHD stages at ARIC Visit 5 with incident death, eachcardiovascular event, or the composite of all outcomes.

		N	Events	Rate per 100 PY	HR Model 1	P-value	HR Model 2	P-value
Death	No VHD	2,640	420(16%)	2.4(2.2–2.6)	Reference group
	Stage A	2,362	515(22%)	3.3(3.1–3.6)	1.3(1.1–1.4)	<0.001	1.2(1.1–1.4)	0.002
	Stage B	1,010	306(30%)	4.8(4.3–5.3)	1.5(1.3–1.7)	<0.001	1.4(1.2–1.7)	<0.001
	Stage C/D	66	38 (58%)	9.6(7.013.1)	3.0(2.2–4.2)	<0.001	2.3(1.6–3.4)	<0.001
HF	No VHD	2,545	153(6%)	0.9(0.8–1.1)	Reference group
	Stage A	2,219	248(11%)	1.7(1.5–2.0)	1.8(1.5–2.2)	<0.001	1.9(1.5–2.3)	<0.001
	Stage B	909	136(15%)	2.4(2.0–2.9)	2.2(1.7–2.8)	<0.001	2.2(1.7–2.9)	<0.001
	Stage C/D	56	16(29%)	5.4(3.3–8.9)	5.0(3.0–8.4)	<0.001	4.7(2.7–8.4)	<0.001
AF	No VHD	2,495	186(7%)	1.4(1.2–1.6)	Reference group
	Stage A	2,115	238(11%)	2.2(1.9–2.5)	1.5(1.2–1.8)	<0.001	1.4(1.2–1.7)	0.001
	Stage B	864	128(15%)	3.0(2.5–3.6)	1.7(1.4–2.2)	<0.001	1.7(1.3–2.1)	<0.001
	Stage C/D	53	12 (23%)	5.2(3.0–9.2)	3.1(1.7–5.6)	<0.001	2.4(1.2–4.7)	0.012
CHD	No VHD	2,359	106(4%)	0.7(0.6–0.8)	Reference group
	Stage A	2,019	123(6%)	0.9(0.8–1.1)	1.4(1.0–1.8)	0.015	1.4(1.0–1.8)	0.023
	Stage B	842	64(8%)	1.2(1.0–1.6)	1.6(1.2–2.2)	0.003	1.6(1.2–2.3)	0.004
	Stage C/D	50	7(14%)	2.6(1.2–5.4)	3.2(1.5–7)	0.003	2.9(1.2 –6.6)	0.013
Stroke	No VHD	2,579	102(4%)	0.6(0.5–0.7)	Reference group
	Stage A	2,276	96(4%)	0.7(0.5–0.8)	1.0(0.8–1.4)	0.842	1.0(0.8–1.4)	0.869
	Stage B	967	50(5%)	0.8(0.6–1.1)	1.2(0.8–1.7)	0.381	1.1(0.8–1.7)	0.468
	Stage C/D	64	2(3%)	0.5(0.1–2.1)	0.7(0.2–3.0)	0.669	0.4(0.1–2.6)	0.318
Composite	No VHD	2,437	413(17%)	3.7(3.3–4.1)	Reference group
	Stage A	2,034	396(19%)	4.5(4.1–4.9)	1.2 (1.0–1.3)	0.041	1.2 (1.0–1.4)	0.038
	Stage B	811	209(26%)	6.3(5.5–7.2)	1.4 (1.2–1.7)	<0.001	1.3 (1.1–1.6)	0.001
	Stage C/D	51	20(39%)	11.4(7.4–17.7)	2.7 (1.7–4.2)	<0.001	2.2 (1.3–3.7)	0.003

Event rates are per 100 person-years. HRs are adjusted. Model 1: adjusted for age, sex, and race; Model 2: adjusted for age, sex, race, Field Center, hypertension, diabetes, prior myocardial infarction, heart failure, body mass index, systolic blood pressure, estimated glomerular filtration rate, and low-density lipoprotein at Visit.

Among 1,010 participants with Stage B VHD, at least two valvular lesions were present in 60% (the additional lesion being Stage A in 49% and Stage B in 11%). Compared to Stage B participants with only 1 Stage B lesion, involvement of a second valvular lesion was associated with a trend toward higher risk for the composite of death, HF, CHD, AF, or stroke (HR 1.34 [95% CI 1.00– 1.80], p= 0.049 in the model adjusted for demographics; HR 1.34[0.98– 1.84], p= 0.069 in the fully adjusted model; Figure S6).

Progression of VHD Stage from Visit 5 to Visit 7

During the 6.6 (IQR 6.1–7.0) years between Visit 5 and Visit 7, 1,223 participants died, and 2,896 (60%) of surviving participants chose to attend Visit 7 and underwent repeat echocardiography. Among those without Visit 7 echocardiography, 25 participants died with VHD-related death certificate ICD codes, 146 were hospitalized with VHD-related ICD codes, and 21 were hospitalized with VHD procedure-related ICD codes (Figure 1). Among the 2896 with repeat echocardiography at Visit 7, mean age was 81± 4 years, 57% were women, and 23% reported Black race. Stage A VHD was present in 19.8%, Stage B in 9.8%, and Stage C/D in 4.1% (Figure 4; Figure S7). After incorporating IPAW to account for Visit 7 non-attendance, the prevalence of no VHD decreased between Visits 5 and 7 (43.2 to 23.8% respectively), as did the prevalence of Stage A VHD (38.6 to 31% respectively), while the prevalence of Stage C/D VHD increased (1.1 to 7.3% respectively) as did the prevalence of valve replacement or repair (1.1 to 2.2% respectively). No major changes were observed in the prevalence of Stage B VHD (16.5 to 15.7% respectively; Figure 4A). Similar reductions in the proportion of participants without VHD stage from Visit 5 to Visit 7 were observed for each valve lesion, and similar increases in the prevalence of Stage C/D VHD were observed for AS and MR (Figure S8, Figure S9).

Figure 4.

Transitions of VHD stages over 6.6 years from ARIC Visit 5 (mean age 76±5 years) to Visit 7 (mean age 81± 4 years) overall (Panel A), by age category (Panel B), and by race and gender groups (Panel C). Sankey diagram in Panel A demonstrates the transition in VHD stage from Visit 5 to Visit 7 (left 2 columns). Right-most column demonstrates prevalence of VHD stages at Visit 7 using IPAW to account for Visit 7 non-attendance. For VHD transitions by age group (Panel B) and race and gender groups (Panel C), Visit 7 prevalence estimates are IPAW-adjusted. For VHD transitions by race and gender groups (Panel C), prevalence estimates are age-adjusted. VHD indicates valvular heart disease.

The magnitude of decline in prevalence of no VHD and increase in Stage C/D prevalence was greater at older age, despite higher mortality between Visits 5 and 7 among older participants (Figure 4B). After accounting for age, changes in prevalence of no VHD and C/D were similar among subgroups defined by gender and race (Figure 4C). A sensitivity analysis limited to 2,896 participants with echocardiography at both Visits 5 and 7 demonstrated similar findings (Figure S10).

Discussion

This analysis is one of the first to quantify the prevalence, prognostic relevance, and progression of ACC/AHA VHD stages in a large, diverse, community-based cohort of persons in late-life. We report three major novel findings. First, less than half of older adults are free of VHD stage, with 39% having Stage A and 17% having Stage B VHD. Second, compared to those free of VHD, a higher VHD stage was associated with a graded increase in risk of incident HF, CHD, AF, and mortality after adjustment for traditional cardiovascular risk factors. Even Stage A VHD was associated with heightened risk for adverse cardiovascular events compared to those free of VHD. Third, VHD stages are progressive, and accelerate in late-life, with a decline in the proportion free of VHD stage from 43 to 24% and an increase in the prevalence of Stage C/D VHD from 1 to 7% over 6 years. These findings, which capture the range of sub-severe VHD in the community, highlight the scope and pace of progression of VHD among older adults. Our findings quantify the substantial burden of VHD in late-life, the prognostic importance of even mild VHD, and the importance of clinical follow-up given the accelerated progression of VHD with older age.

Numerous prior studies have documented the prevalence of individual valvular lesions,^24–30 frequently using clinically referred samples and focusing on greater degrees of lesion severity.^31–34 In contrast, Nkomo et al.’s landmark large pooled analysis of several population-based studies identified moderate or severe VHD in 2.5%, with marked increase in prevalence with age such that 13 % of those ≥75 years old had moderate or severe VHD.¹ More recently, the Oxford Valvular Heart Disease Population Study (OxVALVE) study of 2,500 patients ≥65 years old recruited from primary care clinics demonstrated a prevalence of mild VHD in 44% and moderate or severe VHD in 6.4%.²

In contrast to prior studies, which largely characterize VHD lesions as mild or significant (moderate or severe), our study is the first to our knowledge to implement the ACC/AHA VHD stages framework. While the construct of disease stages has been incorporated into HF guidelines for two decades,³⁵ its incorporation into VHD guidelines is relatively recent.⁴ In addition to emphasizing the progressive nature of valvular lesions, articulation of VHD stages also provides a framework for quantifying the population burden of lesser degrees of valvular dysfunction. Our study now extends upon prior studies by evaluating VHD stages in a diverse, community-based sample of older-adults, defining the association of non-severe VHD with CVD outcomes beyond mortality, and quantifying the progression in VHD stages over 6 years in late-life.

We found that only 43% older adults in a community-based cohort were free of VHD, similar to that observed in the OxVALVE study (49%).² The most common VHD stages in our study were Stage A (at risk) based on valvular deformity, calcification, aortic sclerosis, or mild MR, followed by Stage B (progressive VHD), which included other mild and moderate valvular lesions. The prevalence of mild mitral and aortic valve regurgitation is known to increase with age.^28,36The guideline classification of mild MR as a criterion for Stage A VHD but mild AR as Stage B VHD resulted in the majority of Stage A VHD being due to mild MR or MAC and the majority of Stage B VHD being due to mild AR. The prevalence of the Stage C or D VHD (symptomatic and asymptomatic severe valvular lesions) was 1% at Visit 5 in our study, lower than observed in OxVALVE (6.4%) and the Nkomo et al. study (2.5%). ^1,2 Importantly, the prevalence in our study increased to 7% at Visit 7, more consistent with OxValve and Nkomo et al. The lower prevalence at Visit 5 may therefore be related to healthy attendance bias, leading to underestimation of the true prevalence. Furthermore, the OxVALVE study sampled patients from UK primary care clinics which may have resulted in over-sampling of patients with VHD compared to a community-based cohort. Between-study differences in VHD definition likely also contribute to differences in prevalence estimates. For example, significant AS in the Nkomo et al. study was based on an AVA ≤1.5 cm² assessed variably by each component cohort.^1,2,37

Severe AS and MR are the most common severe valvular lesions in western countries ³⁸ and are associated with risk of all-cause mortality even when asymptomatic.^29,39,40 Consistent with these data, Stage C/D VHD in our study was mainly due to AS and MR and was associated with a markedly increased risk of mortality and incident CVD compared to those free of VHD. Importantly, Stage A and Stage B VHD were also associated with higher concentrations of biomarkers of myocardial stress (NT-prBNP) and injury (hsTn-T), and with a heightened risk mortality, CHD, AF, and HF in a graded fashion even after adjustment for common cardiovascular risk factors. These findings are consistent with prior studies demonstrating the association of mild and moderate AS, aortic sclerosis in the absence of stenosis, mild and moderate MR, and mild degenerative MS with a higher risk of mortality and CVD events such as CHD.^41–46 Notably, the association of Stage A or Stage B AR with CV outcomes was inconsistent suggesting modest risk of CV events associated with AR in late life, although the large majority (92%) of Stage B AR participants had mild – as opposed to moderate – AR in our study. It remains unclear whether Stage A, or even Stage B, VHD is etiologically related to mortality or incident CVD, or is instead a marker of unmeasured cardiovascular risk factors.⁴⁵ Indeed, modest alterations in valve function, such as aortic sclerosis and mild stenosis, are associated with a worse cardiovascular health score after adjusting for demographics.⁴⁷

Limited data exist regarding the progression of VHD, and of VHD stages, particularly in late life when both prevalence and incidence of VHD are greatest. Previous studies have found that progression from aortic sclerosis to clinically significant stenosis occurs in 5.4% of patients over seven years,^48,49 while progression from mild to moderate AS to severe AS occurs in upto 47% of asymptomatic patients over five years. ⁴⁶ Progression in MR severity over time appears more variable, likely because of the diverse etiologies of MR and numerous factors influencing regurgitation severity.⁵⁰ We observed a marked four to seven fold increase in the prevalence of Stage C/D VHD over six years in late-life, from a mean age of 76±5 to 81±4 years. These increases were driven by increases in the prevalence of Stage C/D AS and MR, which demonstrated five- and ten-fold increase respectively. Importantly, the rate of progression in VHD was greater at older ages, despite the greater competing risk of death. This accelerated progression highlights the importance of surveillance for VHD and of developing approaches targeting prevention and/or mitigation of VHD in late life.

This study has several limitations. Non-attendance of surviving ARIC participants at Visit 5 may introduce healthy selection bias and limit generalizability. A sensitivity analysis incorporating IPAW demonstrated similar results. Non-attendance at Visit 7 (30%) was non-random and may therefore introduce attrition bias and underestimation of the prevalence of Stage C/D VHD at Visit 7. We therefore performed analyses incorporating IPAW to account for Visit 7 non-attendance. Furthermore, among Visit 7 non-attendees, we assessed death and hospitalization ICD codes between Visits 5 and 7 to detect clinical progression. The operational definition of AR and MS was based primarily on qualitative assessment by COCATS level III trained cardiologists using ASE criteria. Proximal Isovelocity Surface Area (PISA)-based measures of MR severity were not available for the classification of progressive and severe MR. Information regarding the type and cause of MR was also not available. Vena contracta was not available for the classification of AR stage. The calculated pressure half-time method may have limited accuracy in the setting of significant mitral annular calcification. Furthermore, the low prevalence of MS limited our ability to assess the association of MS stage with clinical outcomes. Details of the primary indication and the type of AV and MV interventions were not available. Data on tricuspid valve structure and function were not available. Finally, residual confounding of the association of VHD stage and risk of cardiovascular outcomes cannot be excluded.

Conclusions

In a diverse community-based cohort of older adults, subclinical VHD is common, with 39% at risk of VHD (Stage A) and 17% with progressive VHD (Stage B). Stage A and Stage B VHD are associated with a heightened risk of incident cardiovascular events independent of traditional cardiovascular risk factors. VHD stages progress over six years in late life, with a several-fold increase in the prevalence of severe VHD (Stage C/D). These findings clarify the burden of VHD in late-life and highlight the public health importance of interventions to mitigate VHD progression.

Clinical Perspective.

What is new:

• In a diverse community-based cohort of older adults, sub-severe VHD is common.

• Stage A and Stage B VHD are associated with a heightened risk of incident cardiovascular events independent of traditional cardiovascular risk factors.

• VHD stages progress over 6 years in late life in the community, with considerable increases in Stage C/D prevalence.

Clinical implications:

• These findings quantify the substantial burden of VHD in late life, and highlight the importance of clinical follow-up given the accelerated progression of VHD in older age.

• Public health interventions are required to mitigate this burden.

Non-standard Abbreviations and Acronyms

ACC/AHA

American College of Cardiology /American Heart Association

AF

Atrial fibrillation

AR

Aortic regurgitation

AS

Aortic stenosis

ASE

American Society of Echocardiography

AV

Aortic valve

AVA

Aortic valve area

CHD

Coronary heart disease

CKD-EPI

Chronic Kidney Disease-Epidemiology Collaboration

COCATS

Core Cardiology Training Symposium

CVD

Cardiovascular diseases

eGFR

Estimated glomerular filtration rate estimated

HF

Heart failure

hs-TnT

High sensitive Troponin T

ICD

International Classification of Diseases

IPAW

Inverse probability of attrition weights

LAA

Left atrial area

LDL

Low-density lipoprotein

MI

Myocardial infarction

MR

Mitral regurgitation

MRJA

Mitral regurgitation jet area

MS

Mitral Stenosis

MVA

Mitral valve area

NT-proBNP

N-terminal pro-brain natriuretic peptide

OxVALVE

Oxford Valvular Heart Disease Population Study

PISA

Proximal Isovelocity Surface Area

VHD

Valvular heart disease

Vmax

Aortic valve peak jet velocity

Access to Data and Data Analysis

Dr. AM Shah had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.