Abstract
Background
The natural history of left ventricular hypertrophy (LVH) - an important risk factor for heart failure (HF) - is heterogeneous.
Objectives
We hypothesized that biomarkers of subclinical myocardial injury (high sensitive cardiac troponin T [hs cTnT]) and hemodynamic stress (NT-proBNP) would differentiate HF risk among older adults with LVH.
Methods
NT-proBNP and hs cTnT were measured at baseline and after 2–3 years in older adults without prior HF or MI in the Cardiovascular Health Study. LVH and LV ejection fraction (EF) were determined by echocardiography. HF events were adjudicated over a median 13.1 years and classified as preserved or reduced LVEF (HFpEF or HFrEF). Adjusted risk of HF by LVH and biomarker tertiles, and by LVH and longitudinal increase in each biomarker, was estimated using Cox regression.
Results
Prevalence of LVH was 12.5% among 2,347 participants with complete measures. Adjusted risk of HF (N=643 events) was roughly 3.8-fold higher among participants with LVH and in the highest biomarker tertile, compared to those with low biomarker levels without LVH (NT-proBNP: HR=3.78 [95% CI: 2.78, 5.15]; hs-cTnT: HR=3.86 [2.84, 5.26]). The adjusted risk of HFrEF was 7.8 times higher among those with the highest tertile of hs cTnT and LVH (HR=7.83 [4.43, 13.83]). Those with LVH and longitudinal increases in hs cTnT or NT-proBNP were roughly 3-fold more likely to develop HF – primarily HFrEF - compared to those without LVH and with stable biomarkers.
Conclusions
The combination of LVH with greater hs cTnT or NT-proBNP levels, and their longitudinal increase, identifies older adults at highest risk for symptomatic HF, especially HFrEF. These biomarkers may characterize sub-phenotypes in the transition from LVH to HF and suggest modifiable targets for prevention.
Keywords: natriuretic peptides, heart failure, left ventricular hypertrophy, epidemiology, troponin T
Introduction
Hypertension is present in greater than 70% of older adults (1), and is commonly associated with LVH (2). Although LVH is associated with an increased risk of progression to depressed left ventricular systolic function, HF, and death, the progression to a clinical endpoint is heterogeneous, occurring in only a minority (3,4). In a prior study of middle age adults, we showed that biochemical evidence of myocardial injury (as measured by the hs cTnT assay) or myocardial hemodynamic stress (as measured by NT-proBNP) identified a “malignant” phenotype of LVH more likely to progress to heart failure or death (5). Currently, routine cardiac imaging to screen for LVH in hypertensive patients is not recommended, and several important questions remain before considering hs cTnT or NT-proBNP as part of a strategy to identify individuals with LVH at high risk for progression to HF (6). First, HF is heterogeneous with a near equivalent incidence of HFpEF and HFrEF (7). Identification of those at highest risk of HFrEF may be particularly advantageous, as specific therapies exist to reduce progression to symptomatic disease (8). However, clinical and echocardiographic characteristics still have a limited ability to differentiate who will progress to HFrEF versus HFpEF (9,10). Our prior study in middle-aged adults was not able to examine this heterogeneity in HF outcomes, nor whether longitudinal changes in cardiac biomarkers may further modify the risk associated with LVH.
The primary objectives of this study were to: 1) Determine if our prior findings in middle-age adults with LVH would be applicable to older adults, and whether there were differential associations with HFrEF vs. HFpEF. Older adults have a markedly higher incidence of HF – especially HFpEF - compared to younger adults, but also greater comorbidities which can confound the interpretation of cardiac-specific biomarkers; and 2) determine whether longitudinal changes in NT-proBNP and/or hs cTnT in those with LVH are associated with the preferential development of HFrEF rather than HFpEF.
Methods
Study Participants
The CHS is a prospective observational study of cardiovascular risk factors in older adults. Detailed descriptions of the methods have been described (11). Study participants included community-dwelling adults ≥65 years enrolled at 4 participating centers. Participants (N=5201) initially enrolled in 1989–90, and an African-American supplemental cohort (N=687) enrolled in 1992–93. For the present analysis, we excluded participants with a prior history of HF, myocardial infarction, or estimated GFR<30 cc/min/1.73m2 at the time of the initial echocardiogram (see below).
The CHS was approved by the IRBs of the University of Washington and participating centers. All participants gave written informed consent. The present analysis was approved by the IRB of the University of Maryland.
Echocardiography
The methods for echocardiographic assessment have been published previously (12). Briefly, two dimensional echocardiography was performed in 1989–90 (main cohort only) and again among both cohorts in 1994–1995. M-Mode measurement of LVM was performed using the method of Deveraux et al (13). LVM could not be estimated in approximately 34% of the main cohort, who were more likely to be older, Caucasian, male, of greater height and weight, and to have hypertension, diabetes, and coronary disease (12). Expected LVM was calculated based on normative equations from CHS participants with neither clinical heart disease nor hypertension; left ventricular hypertrophy was defined as an observed/expected LVM>1.45 (12,14). LVMI was calculated as LVM divided by body surface area. For analyses of baseline biomarkers, LVM measured at baseline (main cohort) or 1994–95 (supplemental cohort) was used as the primary predictor variable; for analyses of change in biomarkers, LV mass measured in 1994–95 in both cohorts was used. LVEF was defined as abnormal if visually interpreted as < 45%. RWT was computed as previously described as (2*Posterior wall thickness)/(End-diastolic LV diameter) (15). Eccentric LVH was defined as LVH with RWT≤0.42, and concentric LVH with RWT>0.42 (15).
Biomarker measurement
NT-proBNP and hs cTnT were measured in serum samples collected at baseline and again after 3 years (main cohort) or 2 years (supplemental cohort) and stored at −70°C to−80°C. NT-proBNP and cTnT were measured on the Elecsys 2010 analyzer (Roche Diagnostics, Indianapolis, Indiana), as previously reported (16,17). The performance characteristics of both assays have been described previously (18).
Primary Outcome
The primary outcome was incident HF, ascertained by participant interview at semiannual study visits, medical record review and examination of Medicare claims data and confirmed by expert adjudication panel as described previously (19). A HF event was confirmed if a physician diagnosis was present along with documentation in the medical record of a constellation of symptoms and physical signs, supporting clinical findings, or a medical therapy for HF. Events were characterized as HFpEF (LVEF≥45%) or HFrEF (LVEF<45%) based on clinical echocardiograms or other cardiac imaging performed within 30 days of the HF event (14).
Statistical Analyses
Participants were divided into age- and sex-specific tertiles of each biomarker, and differences across these tertiles were compared separately for those with and without LVH, using ANOVA for continuous variables and Cuzik’s Score test for binary variables. 895 (38%) of participants had undetectable hs cTnT below level of blank (<3 ng/L) and were all placed in the first tertile with an imputed value of 2.99 ng/L. Cumulative rates of HF among subjects stratified by LVH and biomarker categories were compared with the log-rank test. Cox proportional hazards models were used to estimate the joint association of LVH and biomarker levels with incident heart failure, adjusting for potential confounding factors selected a priori, as defined in Tables 2 & 3, below. The method of Breslow was used to handle tied events (20). Joint associations were estimated using LVH*biomarker interaction terms in adjusted models. Similar analyses were performed using LVMI categories in place of LVH. We estimated improvements in reclassification and discrimination of 10-year HF risk among those with LVH from the addition of each biomarker measurement to traditional risk factors with the net reclassification improvement (NRI) and integrated discrimination improvement (IDI) statistics (21). Consistent with recent recommendations (22), we used the category-less form of the NRI, as there are no consensus thresholds for classifying HF risk among those with LVH. Bootstrapping was used to estimate 95% confidence intervals of each NRI.
Table 2.
Risk of Incident HF, by LVH and initial biomarker level
| Hazard Ratios (95% CI) | |||
|---|---|---|---|
| LVH by echo |
Tertile of NT-proBNP |
Unadjusted | Risk-factor adjusted* |
| None | 1 | 1.0 | 1.0 |
| 2 | 1.22 (0.96, 1.50) | 1.22 (0.97, 1.52) | |
| 3 | 2.03 (1.64, 2.50) | 1.94 (1.56, 2.41) | |
| Yes | 1 | 1.87 (1.20, 2.91) | 1.71 (1.10, 2.67) |
| 2 | 2.52 (1.69, 3.76) | 2.07 (1.38, 3.10) | |
| 3 | 4.42 (3.28, 5.94) | 3.78 (2.78, 5.15) | |
| LVH by echo |
Tertile of hs cTnT |
Unadjusted | Risk-factor adjusted* |
| None | 1 | 1.0 | 1.0 |
| 2 | 1.69 (1.36, 210) | 1.36 (1.09, 1.69) | |
| 3 | 2.75 (2.23, 3.38) | 2.07 (1.67, 2.56) | |
| Yes | 1 | 2.62 (1.80, 3.81) | 2.31 (1.58, 3.36) |
| 2 | 2.20 (1.41, 3.44) | 1.70 (1.08, 2.66) | |
| 3 | 5.88 (4.37, 7.90) | 3.86 (2.84, 5.26) | |
Risk-Factor Adjusted: Age, race, gender, smoking, hypertension, diabetes, coronary heart disease, body mass index, eGFR<60 cc/min/1.73m2, LVEF<45%, and relative wall thickness (RWT).
Interaction of LVH with: NTproBNP: p=0.5; hs cTnT: p=0.6
Table 3.
Risk of Heart failure with reduced EF, by LVH and initial biomarker level
| Hazard Ratios (95% CI) | |||
|---|---|---|---|
| LVH by echo |
Tertile of NT-proBNP |
Unadjusted |
Risk-factor adjusted* |
| None | 1 | 1.0 | 1.0 |
| 2 | 1.03 (0.64, 1.65) | 1.00 (0.62, 1.62) | |
| 3 | 1.74 (1.11, 2.21) | 1.66 (1.05, 2.62) | |
| Yes | 1 | 1.03 (0.64, 1.65) | 0.93 (0.28, 3.04) |
| 2 | 3.49 (1.72, 70.6) | 2.92 (1.42, 5.99) | |
| 3 | 6.48 (3.82, 10.97) | 5.06 (2.89, 8.86) | |
| LVH by echo |
Tertile of hs cTnT |
Unadjusted |
Risk-factor adjusted* |
| None | 1 | 1.0 | 1.0 |
| 2 | 2.30 (1.42, 3.73) | 1.77 (1.08, 2.89) | |
| 3 | 3.64 (2.29, 5.80) | 2.62 (1.62, 4.21) | |
| Yes | 1 | 2.70 (1.12, 6.51) | 2.19 (0.90, 5.32) |
| 2 | 3.38 (1.40, 8.14) | 2.65 (1.10, 6.46) | |
| 3 | 12.94 (7.5, 22.23) | 7.83 (4.43, 13.83) | |
Risk-Factor Adjusted: Adjustment covariates same as for Table 2.
Interaction of LVH with: NT-proBNP tertiles: p=0.07; hs-cTnT tertiles: p=0.4
To examine the joint association of LVH and change in biomarkers with incident HF, Cox proportional hazards models were used, with follow-up time defined as time from the 2nd echocardiogram (1994–95). A significant change in biomarkers was defined as: a) an increase in NT-proBNP of >25% to a final level ≥190 pg/mL, or b) an increase in hs cTnT of>50% from baseline. For those participants with an initial hs cTnT below level of blank, a level of 2.99 ng/L was imputed. Changes of this magnitude for cTnT and NT-proBNP have been associated with marked increases in risk of incident HF and cardiovascular death in CHS (16,17). Adjustments were made for baseline biomarker levels and the same confounding factors as described above, measured at the time of the 2nd echocardiogram (except for eGFR, which was measured at the 1992–93 visit). LVH*biomarker change interaction was tested in multivariate models using the likelihood ratio test. Similar analyses were performed using LVMI categories in place of LVH.
Survival analyses were performed for all incident HF, and separately for incident HFrEF and HFpEF. At-risk time was defined as time from the echocardiogram to incident HF, with censoring on death or last observed follow-up; for analyses of HF subtype, participants were additionally censored at time of HF of any other subtype. In sensitivity analyses, we used the Fine-Gray method to model the competing risk of LVH and each biomarker with all-cause mortality. All statistical analyses were performed with Stata/SE 12.1 (College Station, TX).
Results
Study participants
Among 2347 participants included in the baseline biomarker analyses (supplemental figure 1a), 294 (12.5%) had LVH, among whom 210 had eccentric and 84 concentric LVH. Prevalence of LVH across progressive NT-proBNP tertiles was 8.6%, 10.9%, and 18.1% (p<0.001); and across hs cTnT tertiles was 8.5%, 11.9%, and 19.3%, (p<0.001). Hypertension was present in 74% of those with LVH and 55% without.
Table 1 shows baseline clinical and echocardiographic characteristics, stratified by hs cTnT and presence of LVH. Those with greater hs cTnT were older, more likely to be male, have diabetes, abnormal LVEF, an eGFR<60 cc/min/1.73m2, and higher blood pressure and body mass index. Similar trends across hs cTnT were noted for those with and without LVH. Among subjects without LVH, those with higher NT-proBNP were less likely to be African-American and more likely to have coronary heart disease, an eGFR<60 cc/min/1.73m2, an abnormal LVEF, higher blood pressure and body mass index (supplemental table 1). Similar trends, though typically not significant, were observed among those with LVH. Correlations of NT-proBNP (and hs-cTnT with LVMI at baseline were only modest (ρ=0.12 and ρ=0.21, respectively).
Table 1.
Characteristics of study participants at baseline, by LVH and initial hs cTnT (N=2347)
| No LVH | LVH | |||||||
|---|---|---|---|---|---|---|---|---|
| Tertile 1 | Tertile2 | Tertile 3 | Test for trend |
Tertile 1 | Tertile2 | Tertile 3 | Test for trend |
|
| N | 932 | 584 | 535 | 87 | 79 | 128 | ||
| Range (pg/mL)* | ||||||||
| Men | <3.00 – 4.82 | 4.84 – 9.23 | 9.32 – 43.89 | <3.00 – 4.73 | 5.29 – 7.78 | 9.80 – 49.60 | ||
| Women | <3.00 | 3.00 – 6.00 | 6.02 – 36.12 | <3.00 | 3.12 – 5.98 | 6.55 – 71.82 | ||
| Age (years) | 71.2 (4.6) | 72.7 (5.1) | 73.1 (5.6) | <.001 | 71.5 (4.4) | 73.7 (5.8) | 73.8 (5.9) | .006 |
| Male | 247 (26.5%) | 250 (42.8%) | 216 (40.4%) | 24 (27.6%) | 29 (36.7%) | 61 (47.7%) | ||
| African-American | 143 (15.4%) | 74 (12.7%) | 88 (16.5%) | 0.8 | 12 (13.8%) | 8 (10.1%) | 16 (12.5%) | 0.8 |
| Diabetes | 92 (9.8%) | 90 (15.4%) | 119 (22.2%) | <.001 | 12 (13.8%) | 9 (11.4%) | 28 (21.9%) | .01 |
| Coronary Heart Disease | 78 (8.4%) | 65 (11.1%) | 58 (10.8%) | 0.09 | 7 (8.0%) | 15 (20.3%) | 21 (16.4%) | 0.13 |
| Body Mass Index | 26.1 (4.1) | 26.3 (4.4) | 27.0 (4.8) | <.001 | 26.4 (4.4) | 27.4 (4.7) | 27.1 (4.7) | 0.5 |
| SBP | 132.4 (19.8) | 133.6 (20.3) | 140.1 (22.5) | <.001 | 139.2 (22.8) | 143.8 (19.6) | 144.9 (24.0) | .07 |
| DBP | 70.2 (10.3) | 70.4 (10.8) | 71.4 (11.4) | 0.06 | 71.4 (11.6) | 72.2 (11.2) | 70.9 (13.2) | 0.7 |
| HTNive medications | 330 (35.4%) | 245 (42.0%) | 277 (52.0%) | <.001 | 39 (44.8%) | 46 (58.3%) | 76 (59.4%) | .04 |
| Smoking | 0.9 | 0.12 | ||||||
| Current | 112 (12.0%) | 51 (8.7%) | 51 (9.6%) | 13 (13.9%) | 6 (7.6%) | 16 (12.5%) | ||
| Former | 363 (39.0%) | 260 (44.6%) | 208 (39.0%) | 24 (27.6%) | 25 (31.7%) | 55 (43.0%) | ||
| Never | 455 (48.9%) | 272 (46.7%) | 275 (51.5%) | 50 (57.5%) | 48 (60.8%) | 57 (44.5%) | ||
| eGFRMDRD<60 | 132 (14.2%) | 100 (17.1%) | 148 (27.7%) | <001 | 8 (9.2%) | 17 (21.5%) | 41 (32.0%) | <.001 |
| Abnormal LVEF† | 4 (0.4%) | 5 (0.9%) | 7 (1.3%) | .06 | 1 (1.2%) | 2 (2.5%) | 14 (10.9%) | .002 |
| Relative Wall Thickness | 0.34 [0.30, 0.39] | 0.35 [0.30, 0.40] | 0.35 [0.30, 0.40] | 0.003 | 0.35 [0.28, 0.43] | 0.38 [0.31, 0.42] | 0.37 [0.30, 0.43] | 0.06 |
| LVMI (g/m2) | ||||||||
| Male | 82.6 [71.5, 92.9] | 82.7 [68.5, 96.2] | 81.6 [69.8, 95.3] | 0.9 | 129.2 [119.8, 147.5] | 131.7 [126.2, 138.8] | 128.5 [121.4, 152.4 | 0.9 |
| Female | 71.3 [62.4, 80.5] | 72.4 [61.7, 81.7] | 73.8 [64.2, 85.1] | .04 | 106.4 [103.0, 122.4] | 112.8 [103.4, 121.0] | 113.1 [104.7, 123.1] | 0.1 |
Cell values represent mean (SD), N(%), or median[interquartile range].
Biomarker categories are stratified by age and gender; range for age 70–75 years shown.
LVEF<45% on initial echocardiogram.
Association of LVH and baseline cardiac biomarker levels with incident HF
A total of 643 incident HF events occurred during a median 13.1 years (interquartile range; 7.1, 18.0) of follow-up. The rate of incident HF varied markedly between those with and without LVH (p<0.001) and by tertile of NT-proBNP and hs cTnT (p<0.001 for both, Figures 1a and 1b, respectively). Those participants with LVH and in the highest NT-proBNP tertile were more than four times as likely to have incident HF compared with those without LVH and in the lowest NT-proBNP tertile (Table 2: HR=4.42, 95% confidence intervals [CI]: 3.28, 5.94). Adjustment for demographic factors, co-morbidity, RWT and LVEF attenuated this association only modestly (HR=3.78, 95% CI: 2.78, 5.15). In contrast, those with LVH, but in the lowest NT-proBNP tertile were only at 1.71 (95% CI: 1.10, 2.67) times the risk of incident HF compared to those without LVH, after adjustment. Those participants with LVH and the highest tertile of hs cTnT were at nearly 6 times higher risk of HF compared to those without LVH and in the lowest hs-cTnT tertile (Table 2, 5.88, 95% CI: 4.37, 7.90). This association was attenuated only moderately after adjustment for potential confounders (adjusted HR=3.86, 95%CI: 2.84, 5.26).
Figure 1.
Rate of incident heart failure, by LVH and tertile of NT-proBNP (1a) or hs cTnT (1b). Incident HF ascertained as described in the Methods, and expressed as # incident events per 100 person-years. The study sample was divided into age- and sex-specific tertiles of each biomarker, as described in the Methods. Rate of incident HF varied between those with and without LVH (p<.001) and by tertile of either NT-proBNP or hs cTnT (p<.001 for both).
Similar results were observed when LVMI was used in place of LVH (supplemental table 1). Compared to those in the lowest tertiles of LVMI and biomarker levels, those in the highest tertile of LVMI and biomarkers had 3.4 (NT-proBNP) and 3.3 times (hs cTnT) the risk of incident HF, after adjustment for potential confounders. In a competing risks model accounting for all-cause mortality the associations found in Table 2 were only slightly attenuated and all remained significant (supplemental Table 3). Among those with LVH, additional adjustment for residual differences in LVMI did not change the associations of either biomarker with incident HF (Δβ<5% for both markers). Among those with LVH, the addition of either hs cTnT or NTproBNP (as continuous variables) significantly increased model discrimination, and the addition of hs cTnT significantly improved risk reclassification for incident HF at 10 years when added to traditional risk factors, LVEF, and RWT (supplemental table 4).
LVH, cardiac biomarkers, and incidence of HFrEF versus HFpEF
Among incident HF events, 215 (33.4%) had documented preserved EF, 150 (23.3%) had reduced EF, and 278 (43%) had no documented EF at the time of incident HF diagnosis. Among those with incident HFrEF, 37(24.7%) had LVH at baseline. The rate of incident HF with HFrEF differed significantly by tertile of NT-proBNP among those without (p=0.01) and with (p=0.001) LVH (Figure 2a). The absolute difference in HFrEF rates by NT-proBNP tertile was greater among those with LVH. Similar results were noted for tertiles of hs cTnT, with a markedly greater risk of HFrEF among those with LVH and the highest tertile of hs cTnT (Figure 2b). After adjustment for potential confounders, those participants with LVH in the highest tertile of NT-proBNP had a 5-fold greater risk of incident HFrEF (Table 3: HR=5.06, 95% CI: 2.89, 8.86), and those with LVH in the highest tertile of hs cTnT were at 7.8 times the risk of incident HFrEF versus those without LVH in the lowest hs cTnT level tertile (HR=7.83, 95%: 4.43, 13.83).
Figure 2.
Rate of incident Heart Failure with Reduced Ejection Fraction (HFrEF), by LVH and tertile of NT-proBNP (2a) or hs cTnT (2b). HFrEF ascertained as described in Methods as incident HF with LVEF<45% based on clinical echocardiograms or other cardiac imaging performed within 30 days of the HF event, and expressed as # incident events per 100 person-years. The study sample was divided into age- and sex-specific tertiles of each biomarker, as described in the Methods.
Among those with incident HFpEF, 32 (15.0%) had LVH at baseline. Rate of incident HFpEF also differed significantly by tertile of NT-proBNP level among those without (p=0.003) and with (p=0.02) LVH (Figure 3a). After adjustment for potential confounders, those with LVH and in the highest tertile of NT-proBNP were at approximately 3-fold greater risk of HFpEF compared to those without LVH and with the lowest NT-proBNP (HR=3.11, 95% CI: 1.80, 5.37; supplemental table 5). Similar results were observed for hs cTnT and LVH with regards to risk of HFpEF (figure 3b; supplemental table 5).
Figure 3.
Rate of incident Heart Failure with Preserved Ejection Fraction (HFpEF), by LVH and tertile of NT-proBNP (3a) or hs cTnT (3b). HFpEF ascertained as described in Methods as incident HF with LVEF≥45% based on clinical echocardiograms or other cardiac imaging performed within 30 days of the HF event, and expressed as # incident events per 100 person-years. The study sample was divided into age- and sex-specific tertiles of each biomarker, as described in the Methods.
LVH, change in biomarker levels and incident HF
A total of 1474 subjects had complete measures of change in biomarkers, complete LV mass measures on the 1994–95 echocardiogram and were without HF, MI, or eGFR<30 cc/min/1.73m2. Of these, 193 (13.1%) had LVH.
Participants with LVH and a significant increase of either NT-proBNP or hs cTnT level had markedly higher rates of incident HF (supplemental Figures 2a and 2b; p<0.001 for both biomarkers) compared to those without LVH and with stable or declining biomarker levels. After adjustment for baseline NT-proBNP and risk factors, there remained a nearly 3-fold increased risk for incident HF when NT-proBNP level increased in those with LVH (Table 4). In contrast, those with LVH, but without an increasing NT-proBNP level were not at significantly greater risk of incident HF. Compared to those with stable or declining hs cTnT without LVH, those with LVH and an increase in hs cTnT level were at a 3.1-fold greater risk of incident HF, after adjustment for baseline hs cTnT and risk factors. Similar results were observed using LVMI in place of LVH (Supplemental table 6). In sensitivity analyses, we examined whether the relationship between change in biomarkers and incident HF among those with LVH was explained by residual differences in left ventricular mass between those with and without biomarker increases. Among those with LVH, significant increases of NT-proBNP or hs cTnT were associated with markedly greater HF risk even after additional adjustment for LVMI (as a continuous variable).
Table 4.
Risk of Incident HF, by LVH and change in biomarkers (N=1474)
| Hazard Ratios (95% CI) | ||||
|---|---|---|---|---|
| LVH by Echo |
Increase in NT-proBNP |
% of LVH subgroup |
Baseline- adjusted |
Risk-factor adjusted* |
| None | No | 1046 (81.7%) | 1.0 | 1.0 |
| Yes | 235 (18.3%) | 1.51 (1.14, 2.00) | 1.33 (0.99, 1.80) | |
| Yes | No | 129 (66.8%) | 1.39 (0.97, 1.99) | 1.22 (0.83, 1.78) |
| Yes | 64 (33.2%) | 3.56 (2.46, 5.15) | 2.90 (1.98, 4.27) | |
| LVH by Echo |
Increase in hs cTnT |
% of LVH subgroup |
Baseline- adjusted |
Risk-factor adjusted* |
| None | No | 1062 (82.9%) | 1.0 | 1.0 |
| Yes | 219 (17.1%) | 2.15 (1.63, 2.84) | 1.88 (1.40, 2.50) | |
| Yes | No | 144 (74.6%) | 1.71 (1.23, 2.39) | 1.51 (1.06, 2.16) |
| Yes | 49 (25.4%) | 4.27 (2.85, 6.38) | 3.08 (2.03, 4.67) | |
Baseline-adjusted: Adjusted for baseline biomarker concentration.
Risk-Factor Adjusted: Adjusted for baseline biomarker level, age, race, gender, smoking, hypertension, diabetes, coronary heart disease, body mass index, LVEF<45%, eGFR<60 cc/min/1.73m2, and RWT.
Interaction of LVH with: Increase in NT-proBNP: p=.04; Increase in hs-cTnT: p=0.8
An additional analysis was done to determine the risk of HFrEF and HFpEF based on a rise in biomarker level and the presence or absence of LVH. Compared with individuals without LVH and with no increase in NT-proBNP levels, a rise in NT-proBNP among those with LVH was associated an adjusted HR of 3.46 (95% CI: 1.56, 7.65) for HFrEF but no significant increase in the risk for HFpEF (table 5 and supplemental table 7, respectively). Similarly, compared with individuals without LVH and with no increase in hs cTnT, a rise in hs cTnT in the subgroup with LVH was associated with an adjusted HR of 6.95 (95% CI: 3.07, 15.72) for HFrEF but no increase in the risk for HFpEF (table 5 and supplemental table 7, respectively).
Table 5.
Risk of Incident HF with reduced EF, by LVH and change in biomarker levels (N=1474)
| Hazard Ratios (95% CI) | ||||
|---|---|---|---|---|
| LVH by Echo |
Increase in NT-proBNP |
% of LVH subgroup |
Baseline- adjusted |
Risk-factor adjusted |
| None | No | 1046 (81.7%) | 1.0 | 1.0 |
| Yes | 235 (18.3%) | 1.17 (0.60, 2.29) | 1.14 (0.55, 2.35) | |
| Yes | No | 129 (66.8%) | 2.08 (1.07, 4.06) | 1.99 (0.97, 4.08) |
| Yes | 64 (33.2%) | 4.77 (2.36, 9.77) | 3.46 (1.56, 7.65) | |
| LVH by Echo |
Increase in hs cTnT |
% of LVH subgroup |
Baseline- adjusted |
Risk-factor adjusted |
| None | No | 1062 (82.9%) | 1.0 | 1.0 |
| Yes | 219 (17.1%) | 2.65 (1.45, 4.86) | 2.48 (1.29, 4.77) | |
| Yes | No | 144 (74.6%) | 2.87 (1.53, 5.39) | 2.21 (1.08, 4.54) |
| Yes | 49 (25.4%) | 6.94 (3.22, 14.96) | 6.95 (3.07, 15.72) | |
Cell values are hazard ratios (95% CI) from Cox proportional hazards models. Hazard ratios adjusted for: baseline biomarker level, age, race, gender, smoking, hypertension, diabetes, coronary heart disease, body mass index, eGFR<60 cc/min/1.73m2, relative wall thickness and LVEF<45%.
Interaction between LVH and Change in NT-proBNP: p=0.5
Interaction between LVH and Change in hs-cTnT: p=0.7
Discussion
Among community-dwelling older adults without prior HF or MI, the HF risk associated with LVH was heterogeneous and strongly influenced by baseline levels and changes in NT-proBNP and hs cTnT, biomarkers of subclinical hemodynamic stress and myocardial injury, respectively. Unique to this study was our finding that baseline biomarker elevation appeared to associate with increased risk for progression to HFrEF to a greater extent than HFpEF among those with LVH. The stratification of risk for progression to HFrEF was even more powerful when evaluating longitudinal change in cardiac specific biomarker levels. For example, a rise of >50% in hs cTnT level in combination with LVH was associated with a nearly seven-fold adjusted greater risk of HFrEF, while the same combination of both LVH and an increasing hs cTnT conferred no increased risk for HFpEF. By following longitudinal change, each subject can in effect act as their own control, allowing characterization of the dynamic processes that result in progression from asymptomatic structural heart disease to symptomatic HF.
The implications of our findings are potentially two-fold. First, this study provides clinical data to support a recently proposed paradigm that identifies distinct pathophysiologies for HFrEF and HFpEF (23). Second, the results of this study may provide a rationale to develop and test a preventive strategy utilizing cardiac specific biomarkers and cardiac imaging to identify asymptomatic older adults at highest risk for progression to HFrEF. Though often presenting with similar signs and symptoms, there has been debate as to the degree of commonality of pathophysiology between HFrEF and HFpEF (24). Our findings provide support to the contention that if HFpEF is preceded by myocyte cell hypertrophy, it is without cell death, whereas HFrEF, though potentially preceded by hypertrophy, is also associated with progressive myocyte death and increased wall stress (23,25). This hypothesis is supported by the differences in HF prediction associated with longitudinal change in biomarker levels, which may reflect not only the background milieu of cardiovascular risk factors, but also the pace of asymptomatic myocyte loss and increasing wall stress.
We have previously shown that changes in NT-proBNP and hs cTnT were associated with incident HF and cardiovascular death (16,17). Supporting the concept that myocyte loss – reflected in increases of these biomarkers - is critical to the progression of symptomatic HF, we also demonstrated in older adults with initial low levels of hs cTnT and NT-proBNP and a normal LVEF, that a rise in one or both biomarkers was associated with an increased incidence of progression to asymptomatic reduced LVEF (26). Histologic findings from myocardial biopsies also support evidence of greater myocyte cell loss in those with HFrEF compared to HFpEF (27). In contrast, HFpEF was associated with greater myocyte hypertrophy versus HFrEF irrespective of the extent of collagen deposition (27).
LVH subtype and HF risk
LVH is a well-known structural intermediary in the progression of hypertension to HF.(3,14) However, the progression of LVH to either an abnormal LVEF or symptomatic HF is heterogeneous and cannot be explained on the basis of hypertension alone (3). In other cohorts, dividing LVH into concentric versus eccentric subtypes only moderately differentiated participants at increased risk of HFpEF versus HFrEF (10). This lack of prognostic utility may be secondary to the current two-tiered classification of LVH, which does not account for the presence or absence of LV dilation (28). In the current study, we did not find that NT-proBNP nor hs cTnT were associated with RWT or greater LV mass in those with LVH, nor was there a difference in HFrEF or HFpEF risk by LVH subtype. Overall, our findings suggest that risk stratification among those with LVH may be better achieved by biochemical phenotyping as compared to stratification by relative wall thickness.
Clinical Implications
Current guidelines for hypertension and appropriateness criteria for cardiac imaging do not recommend screening for LVH in hypertensive patients or differentiating treatment based on its presence (6,29). This is in large part based on the low positive predictive value of LVH for HF and no obvious change in treatment strategy based on its identification. However, we and others have previously identified that elevated levels of cardiac specific biomarkers in the presence of LVH stratifies these subjects as particularly high-risk HF (5,30). With extension of this finding in the present study that now identifies HFrEF as a primary sequela of elevated or rising cardiac specific biomarker levels in the presence of LVH in older adults, specific therapies could be considered. In asymptomatic patients with reduced LVEF, ACE inhibitors reduce the progression to symptomatic HF and along with beta blockers remain a class I indication for treatment (8). In patients with LVH and systolic hypertension, an angiotension receptor blockade (ARB) was superior to beta-blockers to prevent a variety of cardiovascular outcomes including HF (4). Further implicating the activation of the renal angiotensin aldosterone system as an upstream mechanism resulting in biochemical measures of myocyte loss and progression to HFrEF are findings from the HOPE study in normotensive vascular risk patients, where a higher dose of ramipril prevented a decrease in LVEF and increase in LV dimensions compared to either a low-dose or placebo (31). We suggest that our findings provide a basis for identifying older adults who could be targeted for HF prevention with renin-angiotensin-aldosterone system (RAAS) antagonism irrespective of whether they have hypertension. Furthermore, even many patients with hypertension may not be treated with an ACE inhibitor or angiotensin receptor blocker (ARB), and if they are treated, the doses can be low and there may be benefit to upward titration.
For example, testing a biomarker strategy to modify care, the STOP-HF trial found intensifying management in primary care patients with at least one cardiovascular risk factor based on mild elevations in BNP resulted in a trend towards reduced new-onset HF.(32) These results were in large part driven by differences in the utilization of ACE inhibitors and ARBs in patients with a known BNP > 50 pg/mL, compared to the control group with a BNP value > 50 pg/mL where the values were unknown to the clinician or patient. Further refining this strategy by evaluating for LVH and measuring NT-proBNP or hs cTnT could identify a high-risk cohort for progression to HFrEF in which specific therapies may be beneficial. A pilot and feasibility study of this approach for primary HF prevention is currently being developed. Lifestyle interventions may also be efficacious, as we have shown in a randomized pilot study that a year of physical activity in previously sedentary older adults significantly blunts a rise in hs cTnT level (33). Greater attention to medical and lifestyle interventions could reduce progression to symptomatic HF in this high-risk cohort with LVH and elevated or rising cardiac specific biomarker levels.
Limitations
LV mass measures were missing in approximately 1/3rd of participants, with missing measures more likely among older male subjects and those with CV risk factors (12). This differential lack of LV mass data may have led to biased estimates of the association with incident HF. However, the fact that these associations persisted after adjustment for demographics and risk factors, and are consistent with our prior findings in younger adults with LVH, suggests these associations are robust. Biomarkers were missing in an additional 25% of participants; as previously reported (16,17), those with complete biomarker measurements differed modestly from those with missing measurements, which could have also introduced bias. We only measured 2 biomarkers, and other biomarkers may have better prognostic utility for HFpEF. Lastly, measures of LVEF at incident HF were incomplete and were not adjudicated by a core echocardiography laboratory, which may have biased the results of associations with HF subtype. However, based on the large number of events with point-of-care echocardiograms, it is unlikely that the robust differences in the prediction of HFrEF versus HFpEF based on biomarkers and LVH would be nullified. Finally, no statistical adjustments to the type I error rate were made for multiple testing, and we cannot exclude a false positive finding.
Conclusion
LVH, as measured by echocardiography, was present in a substantial minority of older adults, particularly in those with elevated levels of NT-proBNP and hs cTnT. The presence of LVH and elevated or rising biomarker levels, independent of risk factors and subclassification of LVH, identified participants at high-risk for new-onset HF, particularly HFrEF. These findings identify a cohort of community dwelling individuals who may ultimately benefit from careful follow-up and consideration of specific medical and lifestyle interventions to prevent progression to symptomatic HF.
Competency in medical knowledge
Left ventricular hypertrophy has long been an ECG and imaging biomarker of risk for incident heart failure and cardiovascular death. Recently, baseline levels and upward trajectories of highly sensitive blood-based cardiac specific biomarkers of cardiac injury (Troponin T) and strain (NT-proBNP) have been shown to predict similar outcomes. The combination of either biomarker with imaging evidence of left ventricular hypertrophy identifies older adults at particularly high risk for developing heart failure with reduced ejection fraction suggesting the cardiac specific biomarkers identify a "malignant" subtype of hypertrophy.
Translational outlook
Left ventricular hypertrophy is a heterogeneous process at a cellular level. Mildly elevated levels of biomarkers of cardiac injury and strain likely identify ongoing subclinical myocyte cell death, potentially through a process of apoptosis rather than ischemic necrosis. Further studies are needed to determine if this subclinical process can be interrupted through pharmacological or lifestyle interventions.
Supplementary Material
Acknowledgments
Funding: This work was supported by contracts HHSN268201200036C, HHSN268200800007C, N01 HC55222, N01HC85079, N01HC85080, N01HC85081, N01HC85082, N01HC85083, N01HC85086, and grant HL080295 from the NHLBI, with additional contribution from the NINDS. Additional support was provided by AG023629 from the NIA. A full list of principal CHS investigators and institutions can be found at CHS-NHLBI.org. Funding for measurement of NT-proBNP and hs cTnT was provided by investigator-initiated grants from Roche Diagnostics Corporation. Dr. Sorkin is funded by NIDDK P30 DK072488, NIA P30 AG028747, and the Baltimore VA GRECC.
Relationships with Industry: Drs. Seliger, deFilippi, and Christenson have received grant support from Roche Diagnostics. Dr. deLemos has received grant support and consulting income from Roche Diagnostics and Abbott Diagnostics. Drs. Seliger, deFilippi, Christenson and deLemos have a patent pending relating to combined LVH and cardiac biomarkers for HF risk stratification
Abbreviations
- hs cTnT
high-sensitivity cardiac troponin T
- NT-proBNP
amino-terminal pro-B-type natriuretic peptide
- LVH
Left Ventricular Hypertrophy
- HF
Heart Failure
- HFrEF
Heart Failure with reduced Ejection Fraction
- HFpEF
Heart Failure with preserved Ejection Fraction
- CHS
Cardiovascular Health Study
- LVM
Left Ventricular Mass
- LVMI
Left Ventricular Mass Index
- RWT
Relative Wall thickness
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Go AS, Mozaffarian D, Roger VL, et al. Heart Disease and Stroke Statistics—2014 Update: A Report From the American Heart Association. Circulation. 2014;129:e28–e292. doi: 10.1161/01.cir.0000441139.02102.80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Lieb W, Xanthakis V, Sullivan LM, et al. Longitudinal tracking of left ventricular mass over the adult life course: clinical correlates of short- and long-term change in the framingham offspring study. Circulation. 2009;119:3085–3092. doi: 10.1161/CIRCULATIONAHA.108.824243. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Drazner MH. The progression of hypertensive heart disease. Circulation. 2011;123:327–334. doi: 10.1161/CIRCULATIONAHA.108.845792. [DOI] [PubMed] [Google Scholar]
- 4.Kjeldsen SE, Dahlof B, Devereux RB, et al. Effects of losartan on cardiovascular morbidity and mortality in patients with isolated systolic hypertension and left ventricular hypertrophy: a Losartan Intervention for Endpoint Reduction (LIFE) substudy. JAMA. 2002;288:1491–1498. doi: 10.1001/jama.288.12.1491. [DOI] [PubMed] [Google Scholar]
- 5.Neeland IJ, Drazner MH, Berry JD, et al. Biomarkers of chronic cardiac injury and hemodynamic stress identify a malignant phenotype of left ventricular hypertrophy in the general population. J Am Coll Cardiol. 2013;61:187–195. doi: 10.1016/j.jacc.2012.10.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Douglas PS, Garcia MJ, Haines DE, et al. 2011 Appropriate Use Criteria for Echocardiography. J Am Soc Echocardiogr. 2011;24:229–267. doi: 10.1016/j.echo.2010.12.008. [DOI] [PubMed] [Google Scholar]
- 7.Owan TE, Hodge DO, Herges RM, et al. Trends in prevalence and outcome of heart failure with preserved ejection fraction. N Engl J Med. 2006;355:251–259. doi: 10.1056/NEJMoa052256. [DOI] [PubMed] [Google Scholar]
- 8.Yancy CW, Jessup M, Bozkurt B, et al. 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on practice guidelines. Circulation. 2013;128:e240–e319. doi: 10.1161/CIR.0b013e31829e8776. [DOI] [PubMed] [Google Scholar]
- 9.De Keulenaer GW, Brutsaert DL. Systolic and Diastolic Heart Failure Are Overlapping Phenotypes Within the Heart Failure Spectrum. Circulation. 2011;123:1996–2005. doi: 10.1161/CIRCULATIONAHA.110.981431. [DOI] [PubMed] [Google Scholar]
- 10.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;6:279–286. doi: 10.1161/CIRCHEARTFAILURE.112.972828. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Fried LP, Borhani NO, Enright P, et al. The Cardiovascular Health Study: design and rationale. Ann Epidemiol. 1991;1:263–276. doi: 10.1016/1047-2797(91)90005-w. [DOI] [PubMed] [Google Scholar]
- 12.Gardin JM, Siscovick D, Anton-Culver H, et al. Sex, age, and disease affect echocardiographic left ventricular mass and systolic function in the free-living elderly. The Cardiovascular Health Study. Circulation. 1995;91:1739–1748. doi: 10.1161/01.cir.91.6.1739. [DOI] [PubMed] [Google Scholar]
- 13.Devereux RB, Alonso DR, Lutas EM, et al. Echocardiographic assessment of left ventricular hypertrophy: comparison to necropsy findings. Am J Cardiol. 1986;57:450–458. doi: 10.1016/0002-9149(86)90771-x. [DOI] [PubMed] [Google Scholar]
- 14.Aurigemma GP, Gottdiener JS, Shemanski L, et al. Predictive value of systolic and diastolic function for incident congestive heart failure in the elderly: the cardiovascular health study. J Am Coll Cardiol. 2001;37:1042–1048. doi: 10.1016/s0735-1097(01)01110-x. [DOI] [PubMed] [Google Scholar]
- 15.Lang RM, Bierig M, Devereux RB, et al. Recommendations for chamber quantification. Eur J Echocardiol. 2006;7:79–108. doi: 10.1016/j.euje.2005.12.014. [DOI] [PubMed] [Google Scholar]
- 16.deFilippi CR, Christenson RH, Gottdiener JS, et al. Dynamic cardiovascular risk assessment in elderly people. The role of repeated N-terminal pro-B-type natriuretic peptide testing. J Am Coll Cardiol. 2010;55:441–450. doi: 10.1016/j.jacc.2009.07.069. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.deFilippi CR, de Lemos JA, Christenson RH, et al. Association of serial measures of cardiac troponin T using a sensitive assay with incident heart failure and cardiovascular mortality in older adults. JAMA. 2010;304:2494–2502. doi: 10.1001/jama.2010.1708. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Giannitsis E, Kurz K, Hallermayer K, et al. Analytical validation of a high-sensitivity cardiac troponin T assay. Clin Chem. 2010;56:254–261. doi: 10.1373/clinchem.2009.132654. [DOI] [PubMed] [Google Scholar]
- 19.Ives DG, Fitzpatrick AL, Bild DE, et al. Surveillance and ascertainment of cardiovascular events. The Cardiovascular Health Study. Ann Epidemiol. 1995;5:278–285. doi: 10.1016/1047-2797(94)00093-9. [DOI] [PubMed] [Google Scholar]
- 20.Breslow N. Covariance analysis of censored survival data. Biometrics. 1974;30:89–99. [PubMed] [Google Scholar]
- 21.Pencina MJ, D'Agostino RB, Sr, D'Agostino RB, Jr, et al. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med. 2008;27:157–172. doi: 10.1002/sim.2929. [DOI] [PubMed] [Google Scholar]
- 22.Leening MJVM, Witteman JCM, Pencina MJ, Steyerberg EW. Net Reclassification Improvement: Computation, Interpretation, and Controversies: A Literature Review and Clinician's Guide. Ann Intern Med. 2014;160:122–131. doi: 10.7326/M13-1522. [DOI] [PubMed] [Google Scholar]
- 23.Paulus WJ, Tschope C. A novel paradigm for heart failure with preserved ejection fraction: comorbidities drive myocardial dysfunction and remodeling through coronary microvascular endothelial inflammation. J Am Coll Cardiol. 2013;62:263–271. doi: 10.1016/j.jacc.2013.02.092. [DOI] [PubMed] [Google Scholar]
- 24.Borlaug BA, Redfield MM. Diastolic and systolic heart failure are distinct phenotypes within the heart failure spectrum. Circulation. 2011;123:2006–2013. doi: 10.1161/CIRCULATIONAHA.110.954388. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Gonzalez A, Ravassa S, Beaumont J, et al. New targets to treat the structural remodeling of the myocardium. J Am Coll Cardiol. 2011;58:1833–1843. doi: 10.1016/j.jacc.2011.06.058. [DOI] [PubMed] [Google Scholar]
- 26.Glick D, deFilippi CR, Christenson R, et al. Long-Term Trajectory of Two Unique Cardiac Biomarkers and Subsequent Left Ventricular Structural Pathology and Risk of Incident Heart Failure in Community-Dwelling Older Adults at Low Baseline Risk. JACC: Heart Failure. 2013;1:353–360. doi: 10.1016/j.jchf.2013.04.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.van Heerebeek L, Borbely A, Niessen HW, et al. Myocardial structure and function differ in systolic and diastolic heart failure. Circulation. 2006;113:1966–1973. doi: 10.1161/CIRCULATIONAHA.105.587519. [DOI] [PubMed] [Google Scholar]
- 28.Khouri MG, Peshock RM, Ayers CR, et al. A 4-Tiered Classification of Left Ventricular Hypertrophy Based on Left Ventricular Geometry: The Dallas Heart Study. Circ Cardiovasc Imaging. 2010;3:164–171. doi: 10.1161/CIRCIMAGING.109.883652. [DOI] [PubMed] [Google Scholar]
- 29.James PA, Oparil S, Carter BL, et al. 2014 Evidence-Based Guideline for the Management of High Blood Pressure in Adults: Report From the Panel Members Appointed to the Eighth Joint National Committee (JNC 8) JAMA. 2013 doi: 10.1001/jama.2013.284427. [DOI] [PubMed] [Google Scholar]
- 30.Olsen MH, Wachtell K, Nielsen OW, et al. N-terminal brain natriuretic peptide predicted cardiovascular events stronger than high-sensitivity C-reactive protein in hypertension: a LIFE substudy. J Hyperten. 2006;24:1531–1539. doi: 10.1097/01.hjh.0000239288.10013.04. [DOI] [PubMed] [Google Scholar]
- 31.Lonn E, Shaikholeslami R, Yi Q, et al. Effects of ramipril on left ventricular mass and function in cardiovascular patients with controlled blood pressure and with preserved left ventricular ejection fraction: a substudy of the Heart Outcomes Prevention Evaluation (HOPE) Trial. J Am Coll Cardiol. 2004;43:2200–2206. doi: 10.1016/j.jacc.2003.10.073. [DOI] [PubMed] [Google Scholar]
- 32.Ledwidge M, Gallagher J, Conlon C, et al. Natriuretic peptide-based screening and collaborative care for heart failure: the STOP-HF randomized trial. JAMA. 2013;310:66–74. doi: 10.1001/jama.2013.7588. [DOI] [PubMed] [Google Scholar]
- 33.deFilippi CR, deLemos J, Newman AB, et al. Abstract 16937: Initiation of Moderate Physical Activity Reduces Progression of Subclinical Cardiac Injury in Previously Sedentary Older Adults: Results From a Randomized Pilot Study of Exercise Intervention. Circulation. 2013;128:A16937. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.