Patients with hypertension and left ventricular (LV) hypertrophy commonly have impaired diastolic filling. However, it remains unknown whether changes in LV diastolic filling variables are associated with cardiovascular morbidity and mortality. In this study, 778 patients with hypertension with electrocardiographic LV hypertrophy who underwent echocardiography at baseline and annually thereafter during randomized losartan- or atenolol-based antihypertensive treatment were followed for a mean of 4.6 years. The composite cardiovascular end point was the first occurrence of fatal or nonfatal myocardial infarction, fatal or nonfatal stroke, and cardiovascular mortality. Antihypertensive therapy resulted in an increase in the prevalence of normal transmitral flow pattern from 28% to 46% of patients. Although antihypertensive treatment often resulted in a marked increase in the prevalence of normal mitral valve flow pattern, this was not associated with reduced cardiovascular morbidity and mortality when adjusting for blood pressure, left atrial diameter, LV mass index, and treatment in time-varying Cox analyses. In contrast, lower in-treatment E/A ratios and shorter mitral valve deceleration times were associated with less risk for heart failure. Similarly, normal in-treatment transmitral flow pattern was strongly associated with less risk for heart failure (hazard ratio 0.22, 95% confidence interval 0.05 to 0.98, p = 0.048), even when taking in-treatment left atrial diameter and blood pressure into account. In conclusion, antihypertensive treatment in patients with hypertension with electrocardiographic LV hypertrophy resulted in significant improvement in transmitral flow patterns; this was not associated with reduced cardiovascular morbidity and mortality. However, normal in-treatment LV filling was strongly associated with a reduced risk for hospitalization for heart failure.
The ratio of peak early to late diastolic left ventricular (LV) filling velocity (E/A ratio) is a widely used clinical measure of LV diastolic filling. Abnormal E/A ratios have predicted poor outcomes in patients with hypertension, dilated cardiomyopathy, and myocardial infarctions and in samples of the general population. We have shown that diastolic filling variables are strongly related to LV hypertrophy and relative wall thickness in patients with hypertension and that reduction in LV mass by antihypertensive treatment may improve transmitral flow variables. This might be a result of LV geometric remodeling, reduced myocardial hypertrophy, and alterations of collagen structure. However, it is still unknown whether LV diastolic function at baseline and improvement of transmitral flow with antihypertensive therapy has a favorable impact on cardiovascular morbidity and mortality in treated patients with hypertension with LV hypertrophy. We hypothesized that >4 years of antihypertensive treatment would improve LV diastolic function and that this would be related to reduced cardiovascular morbidity and mortality.
Methods
The main Losartan Intervention for End Point Reduction in Hypertension (LIFE) outcome as well as the complete study protocol with study design, organization, clinical measures, end point definitions, exclusion criteria, basis for choice of comparative agents, statistical considerations, and baseline characteristics have been previously published.
Eligible individuals were 960 patients with stage II or III hypertension enrolled in the LIFE echocardiography substudy at baseline. However, 182 participants in the echocardiography substudy were ineligible for the present study because of an inability to obtain complete baseline measurements of isovolumic relaxation time (IVRT), mitral valve E/A ratio, or mitral valve deceleration time. Compared with those without needed transmitral flow measurements, the present study’s subjects (n = 778) were more often women and were less likely to have diabetes, but they did not differ in age; blood pressure (BP); body mass index; heart rate; history of atrial fibrillation or coronary, cerebral, or peripheral vascular disease; or incidence of composite end points during follow-up (all p = NS, data not shown). The present report uses the echocardiograms at baseline and annual revisits as well as end points collected during 3,618 patient-years of follow-up.
Patients gave informed consent, and ethics committees in participating countries accepted this study. Before enrollment in the study, all patients underwent screening electrocardiography showing LV hypertrophy by gender-adjusted Cornell voltage-duration product ≥2,440 mV × ms and/or Sokolow-Lyon voltage criterion >38 mV. Further inclusion criteria included lack of myocardial infarction or stroke within 6 months, absence of current congestive heart failure, a known LV ejection fraction <40%, significant aortic stenosis, or overt renal insufficiency (serum creatinine >160 μmol/L [1.8 mg/dl]).
Blinded treatment was begun with losartan or atenolol 50 mg. Study therapy was up-titrated by adding hydrochlorothiazide 12.5 mg, followed by losartan or atenolol 100 mg, to reach target BP of 140/90 mm Hg. Further increase in hydrochlorothiazide to 25 mg and antihypertensive drugs other than angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, or β blockers could be initiated.
Echocardiographic procedures and measurements of left atrial diameter and LV mass from 2-dimensional parasternal views have been previously described. Pulsed Doppler recordings of transmitral flow velocities were obtained between the tips of the mitral valve leaflets. Measurements included peak early LV filling velocity, peak atrial filling velocity, E/A ratio and acceleration, deceleration, and pressure halftimes of early diastolic transmitral flow, and the proportion of the time-velocity integral that occurred in the first third or half of diastole or in response to atrial systole. Pulsed Doppler recording obtained from the LV outflow tract were used to measure IVRT from the closure spike of the aortic valve to the onset of mitral inflow. Measurements for up to 3 cardiac cycles were averaged.
E/A ratio <0.8 was regarded as low in the age group studied in LIFE, and E/A ratio ≥1.5 was considered pathologically high. IVRT ≥100 ms was considered prolonged. Deceleration time <150 ms was regarded as short, and deceleration time >250 ms was considered prolonged. LV filling patterns were classified using a modification of the approach of Appleton et al: abnormal relaxation was identified by low E/A ratio (<1.0), prolonged IVRT (≥100 ms), and mitral valve relaxation time <250 ms (n = 516); a restrictive pattern was identified by E/A ratio ≥1.5 and short IVRT (<100 ms) (n = 34); a pseudonormal filling pattern was classified as the combination of prolonged IVRT (≥100 ms) and normal E/A ratio (n = 84); and LV filling was considered normal with IVRT <100 ms, deceleration time ≥150 and <250 ms, and E/A ratio ≥1.0 and <1.5 (n = 21). Twenty-three patients were left unclassified by this approach, all with low E/A ratios (<1.0), short deceleration times (<150 ms), and normal IVRTs (<100 ms). Compared to patients with classifiable diastolic filling patterns, the remaining had slightly higher systolic BP and higher urine albumin/creatinine ratios but did not differ in age, gender, diastolic BP, body mass index, heart rate, history of diabetes, atrial fibrillation, or coronary, cerebral, or peripheral vascular disease. Furthermore, no events occurred in this group (data not shown).
The primary composite end point, consisting of the first occurrence of cardiovascular death, nonfatal stroke, or nonfatal myocardial infarction, occurred in 77 patients. Additional end points included the first occurrence of each component of the composite end point, whether or not preceded by another component of the primary end point, including 46 fatal or nonfatal strokes and 28 fatal or nonfatal myocardial infarctions, 23 hospitalization for heart failure, and 19 cardiovascular deaths. All end points were reported by investigators, with source data verified by independent monitors and adjudicated by an independent committee on the basis of definitions provided in a predefined end point manual.
SPSS version 17.0 (SPSS, Inc., Chicago, Illinois) was used for statistical analysis. Results are expressed as mean ± SD or as frequencies expressed as percentages. Comparisons among multiple observations were performed by using mixed-effects linear models to account for within-subject correlations. LV diastolic filling variables were divided into 3 groups: IVRT (<100, ≥100 to <125, and ≥125 ms, reflecting normal, mildly to moderately abnormal, and severely abnormal values), mitral valve E/A ratio (<0.8, ≥0.8 to <1.5, and ≥1.5, representing low, intermediate, and high ranges shown to be associated with prognosis in a previous study), and relatively low, normal, and prolonged mitral valve deceleration times (<150, ≥150 to 250, and ≥250 ms, respectively) to examine relations to end points using Cox proportional-hazards models.
Changes in transmitral flow variables were calculated as the difference between measurements on the baseline and last echocardiograms or the echocardiogram before the occurrence of an end point (e.g., composite end point, cardiovascular mortality, fatal and nonfatal myocardial infarction, or fatal and nonfatal stroke) in individual patients. Cox proportional-hazards regression analysis was used to assess the association of baseline transmitral flow variables with an end point. Analyses were adjusted using baseline systolic BP, heart rate, left atrial diameter, and LV mass index and randomized treatment as standard covariates.
To determine whether in-treatment transmitral flow variables predicted clinical end points independent of change in BP heart rate, left atrial diameter, and LV hypertrophy, hazard ratios for in-treatment transmitral flow variables were adjusted by in-treatment systolic BP, heart rate, left atrial diameter, and LV mass index using time-varying Cox proportional-hazards models. The time-varying covariate value associated with an event was the most recent measurement before the event.
All end points were analyzed using the intention-to-treat approach; all randomized patients were included in their randomized treatment group, and all available follow-up data were included from randomization through the study termination date. Patients with multiple end points were counted as having had events in all relevant end point analyses, but only the first event in a specific category counted in any individual analysis. Tests were performed at 2-sided 5% significance levels.
Results
Descriptive data of the whole LIFE echocardiography sample and LV diastolic function at baseline and after 1 year of treatment in the LIFE echocardiography substudy have been reported elsewhere. Baseline characteristics of the present study sample are listed in Table 1 . Patients were on average treated with 81.8 ± 24.6 mg of losartan and 77.0 ± 27.0 mg of atenolol. In addition, hydrochlorothiazide was added in 90.8% of patients randomized to losartan (receiving an average of 17.9 ± 9.6 mg hydrochlorothiazide) and in 85.7% of patients randomized to atenolol (receiving an average of 17.8 ± 11.8 mg hydrochlorothiazide). Changes during treatment are listed in Table 2 and shown in Figure 1 .
| Variable | Value |
|---|---|
| Age (years) | 66 ± 7 |
| Women | 341 (44%) |
| Systolic blood pressure (mm Hg) | 174 ± 14 |
| Diastolic blood pressure (mm Hg) | 98 ± 9 |
| Body mass index (kg/m 2 ) | 27.3 ± 4.5 |
| Heart rate (beats/min) | 67 ± 12 |
| Framingham risk score (5-year risk) | 22.6 ± 9.3 |
| Diabetes mellitus | 78 (10%) |
| Coronary heart disease | 118 (15.2%) |
| Cerebral vascular disease | 62 (8%) |
| Peripheral vascular disease | 43 (6%) |
| Isolated systolic hypertension | 111 (14%) |
| History of atrial fibrillation | 24 (3%) |
| Potassium (mmol/L) | 4.2 ± 0.4 |
| Sodium (mmol/L) | 140.2 ± 2.4 |
| Serum creatinine (mmol/L) | 87.3 ± 19.1 |
| Cholesterol (mmol/L) | 6.0 ± 1.1 |
| High-density lipoprotein cholesterol (mmol/L) | 1.5 ± 0.5 |
| Urine albumin/creatinine ratio (mg/mmol) | 6.5 ± 21.9 |
| Variable | Baseline | 12 mo | 24 mo | 36 mo | 48 mo | 60 mo | Last | p Value ⁎ |
|---|---|---|---|---|---|---|---|---|
| n | 778 | 653 | 626 | 564 | 529 | 256 | 758 | <0.001 |
| Systolic blood pressure (mm Hg) | 174 ± 14 | 149 ± 17 | 146 ± 16 | 144 ± 16 | 144 ± 16 | 142 ± 15 | 144 ± 16 | <0.001 |
| Diastolic blood pressure (mm Hg) | 98 ± 9 | 86 ± 10 | 84 ± 9 | 83 ± 9 | 82 ± 9 | 81 ± 9 | 82 ± 10 | <0.001 |
| Heart rate (beats/min) | 67 ± 12 | 63 ± 12 | 62 ± 12 | 63 ± 15 | 63 ± 11 | 67 ± 47 | 63 ± 12 | <0.001 |
| Body mass index (kg/m 2 ) | 27.3 ± 4 | 27.1 ± 3.9 | 27.3 ± 3.9 | 27.3 ± 4.2 | 27.2 ± 4.1 | 27.1 ± 4.2 | 27.5 ± 4.8 | 0.689 |
| Isovolumic relaxation time (ms) | 115 ± 23 | 104 ± 21 | 99 ± 18 | 99 ± 19 | 97 ± 22 | 95 ± 21 | 96 ± 22 | <0.001 |
| Mitral valve E/A ratio | 0.87 ± 0.38 | 0.93 ± 0.33 | 0.92 ± 0.28 | 0.90 ± 0.27 | 0.94 ± 0.43 | 0.94 ± 0.68 | 0.96 ± 0.57 | <0.001 |
| Mitral valve deceleration time (ms) | 214 ± 64 | 229 ± 65 | 238 ± 60 | 252 ± 65 | 247 ± 69 | 251 ± 70 | 249 ± 69 | <0.001 |
| Atrial filling fraction (%) | 42 ± 11 | 38 ± 11 | 38 ± 10 | 38 ± 9 | 39 ± 10 | 38 ± 10 | 39 ± 11 | <0.001 |
Tertiles of baseline IVRT, E/A ratio, and mitral valve deceleration time and relations to composite end point rate are shown in Figure 2 .
Abnormal transmitral flow patterns at baseline were not associated, compared to the group with apparently normal filling, with higher composite end point rates with abnormal relaxation, pseudonormal filling, or restrictive filling.
Using Cox models, associations of in-treatment IVRT, in-treatment E/A ratio, and in-treatment mitral valve deceleration time with composite end points, cardiovascular mortality, stroke, or myocardial infarction and heart failure are listed in Table 3 . The lack of association of the measurements of diastolic filling with composite end points was not affected by the significant effects of time-varying LV mass index and left atrial diameter (both p <0.05 in all models) or their exclusion from the models (data not shown). In contrast to the in-treatment transmitral flow variables, there were strong associations between reduction in left atrial diameter per SD with reduction in cardiovascular morbidity and mortality ( Table 4 ).
| Variable | HR | 95% CI | p Value |
|---|---|---|---|
| IVRT per SD ⁎ | |||
| Composite end point | 1.14 | 0.92–1.42 | 0.223 |
| Cardiovascular mortality | 0.91 | 0.59–1.40 | 0.661 |
| Stroke | 1.13 | 0.85–1.50 | 0.406 |
| Myocardial infarction | 1.23 | 0.87–1.73 | 0.239 |
| Heart failure | 1.06 | 0.74–1.53 | 0.736 |
| Mitral valve E/A ratio per SD † | |||
| Composite end point | 0.95 | 0.81–1.13 | 0.584 |
| Cardiovascular mortality | 1.06 | 0.88–1.28 | 0.508 |
| Stroke | 1.03 | 0.88–1.19 | 0.746 |
| Myocardial infarction | 0.91 | 0.67–1.24 | 0.559 |
| Heart failure | 1.20 | 1.13–1.29 | <0.001 |
| Mitral valve deceleration time per SD ‡ | |||
| Composite end point | 1.10 | 0.91–1.33 | 0.33 |
| Cardiovascular mortality | 0.95 | 0.64–1.43 | 0.816 |
| Stroke | 0.94 | 0.71–1.22 | 0.627 |
| Myocardial infarction | 1.26 | 0.94–1.69 | 0.123 |
| Heart failure | 1.19 | 1.10–1.29 | <0.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
