The hypothesis that left ventricular hypertrophy (LVH) Regression of Left Ventricular Hypertrophy and Prevention of Stroke in Hypertensive Subjects

Transcription

1 AJH 2006; 19: Heart Regression of Left Ventricular Hypertrophy and Prevention of Stroke in Hypertensive Subjects Paolo Verdecchia, Fabio Angeli, Roberto Gattobigio, Mariagrazia Sardone, Sergio Pede, and Gian Paolo Reboldi Background: Left ventricular hypertrophy (LVH) is a risk marker for stroke and its regression confers protection from stroke. The relationship between serial LVH changes and risk of stroke has never been investigated in a mixed population of hypertensive subjects with and without LVH. Methods: In this study, 880 initially untreated hypertensive subjects (mean age 48 years, office blood pressure (BP) 155/98 mm Hg; 24-h BP 137/87 mm Hg) underwent tests including echocardiography and 24-h ambulatory BP monitoring at entry and after a median of 3.5 years, still in the absence of cardiovascular events. Results: Months or years after the follow-up study, 34 of these subjects developed a first cerebrovascular event (stroke in 21, transient ischemic attack in 13). Event rate ( 100 patients per year) was 0.25 among the subjects who never developed echocardiographic LVH or with regression of LVH, versus 1.16 among the subjects with lack of regression or new development of LVH (log-rank The hypothesis that left ventricular hypertrophy (LVH) is an independent risk marker of the long-term exposure to the combined effects of various risk factors for atherosclerosis 1 is very attractive when it is used to explain the association between left ventricular mass (LVM) and cerebrovascular disease. The heart weight at necropsy is increased in patients with intracerebral hemorrage 2 and an association exists in hypertensive humans between LVM and carotid atherosclerosis. 3 An association between LVH and stroke has been reported in elderly cohorts 4,5 and in hypertensive subjects. 6 However, it is not clear whether serial changes in LVH during treatment are independent determinants of the subsequent risk of stroke. Serial changes in the electrocardiographic test: P.00001). Serial electrocardiogram (ECG) changes failed to define groups at different risk. In a Cox analysis, the risk of cerebrovascular events was 2.8 times higher (95% CI: ) in the subset with lack of regression or new development of LVH than in that with LVH regression or persistently normal LV mass. Such effect was independent of age (P.001) and 24-h systolic BP (P.003). Conclusions: In a mixed hypertensive population with and without LVH at entry, serial changes in the echocardiographic indexes of LVH predict subsequent cerebrovascular events independently of office and ambulatory BP and other individual risk factors. Am J Hypertens 2006; 19: American Journal of Hypertension, Ltd. Key Words: Hypertension, stroke, hypertrophy, prognosis, blood pressure, epidemiology. (ECG) features of LVH predicted the risk of a composite pool of cardiovascular events. 7 Several small studies in hypertensive subjects and a meta-analysis 8 suggest that LVH regression confers protection from a composite of cardiovascular disease. In the Losartan Intervention For End-Point reduction (LIFE) study, carried out in a selected population of hypertensive subjects with LVH at entry, those randomized to losartan showed a greater regression of LVH 9,10 and a lesser risk for stroke 10 than those randomized to atenolol. The present study tested the hypothesis that serial changes in LVH predict the risk of cerebrovascular events in a large hypertensive population of subjects with and without LVH at entry. Received June 23, First decision October 27, Accepted October 27, From the Struttura Complessa di Cardiologia, Hospital R. Silvestrini (PV, FA, RG, MS) Perugia; Dipartimento di Medicina Interna, University of Perugia (GRR); Perugia; and Ospedale N. Melli, San Pietro Vernotico (SP), Italy. This work was supported in part by grants from the Associazione Umbra Cuore e Ipertensione, Perugia, Italy. Address correspondence and reprint requests to Dr. Paolo Verdecchia, Struttura Complessa di Cardiologia, Ospedale R. Silvestrini, Località S. Andrea delle Fratte, Perugia, Italy; verdec@tin.it 2006 by the American Journal of Hypertension, Ltd. Published by Elsevier Inc /06/$32.00 doi: /j.amjhyper

2 494 REGRESSION OF HYPERTROPHY AND STROKE AJH May 2006 VOL. 19, NO. 5 Methods The Progetto Ipertensione Umbria Monitoraggio Ambulatoriale (PIUMA) study is a prospective observational registry of morbidity and mortality in initially untreated subjects with essential hypertension. 6 Entry criteria include an office blood pressure (BP) 140 mm Hg systolic or 90 mm Hg diastolic on at least three visits and absence of secondary causes of hypertension, previous cardiovascular disease, and life-threatening conditions. The BP was measured by a physician with a mercury sphygmomanometer with subjects sitting and relaxed from at least 10 min. Three measurements were averaged for analysis. Systolic and diastolic BP were identified by Korotkoff phases I and V. Twelve-lead ECG was recorded at 25 mm/sec and 1 mv/cm calibration. Subjects with complete right or left bundle branch block, previous myocardial infarction, Wolff-Parkinson-White syndrome, and atrial fibrillation at entry were excluded. None of the subjects was being treated with digitalis. Diagnosis of LVH by ECG was made by using a score recently developed in our laboratory, 11 defined by the presence of either a typical LV strain pattern or a Cornell voltage (sum of R in in a VL Sin V3) 2.0 mv in women or 2.4 mv in men, or both. Ambulatory BP was recorded using an oscillometric device (SpaceLabs 5200, and 90207, SpaceLabs, Redmond, WA) and measurements were automatically taken every 15 minutes throughout the 24 hours. Day and night have been defined using arbitrarily defined fixed time intervals, from 10 AM to 8 PM for day and from midnight to 6 AM for night. Diabetes was diagnosed by fasting plasma glucose of 7.0 mmol/l (126 mg/dl) or higher or current antidiabetic therapy. The M-mode echocardiographic study of the left ventricle was performed under cross-sectional control. Details about reading procedures and reproducibility in our laboratory are reported elsewhere. 12 The LVM was calculated according to Devereux et al 13 and corrected by height at the power of 2.7 to allow for obesity. 14 The LVH was defined by a LVM 51.0 g/height (in meters). 2.7 Follow-Up Follow-up of patients was in charge of family doctors and our hospital staff. Treatment was tailored individually and based on lifestyle and pharmacologic measures. Results of baseline tests were made available to family doctors. It is difficult to establish to what extent the results of some diagnostic tests, particularly 24-h ambulatory BP monitoring and echocardiography, influenced the management of the subjects. Thiazide diuretics, -blockers, angiotensin-converting-enzyme (ACE) inhibitors, angiotensin II (AT-II) antagonists, calcium-channel blockers, and 1 -blockers, alone or combined, were the antihypertensive drugs most frequently used. A comprehensive follow-up examination including standard laboratory tests, 12-lead ECG, echocardiography, and 24-h ambulatory BP monitoring in all subjects was undertaken after a median of 3.5 years. Current antihypertensive treatment was not discontinued and drugs regularly taken at least during the past 3 months were recorded. Protocol for experimental procedures was the same as that in the baseline studies. Notably, none of the patients had developed a cardiovascular morbid event at the time of the follow-up visit. Subsequently to the follow-up visit, periodical contacts with family doctors and phone interviews and clinical visits with patients continued to ascertain the vital status and the occurrence of events. The total duration of follow-up up to event or censoring was up to 18 years (median, 7.2 years). Assessment of End Points Hospital record forms and other source documents of patients who died or suffered a cerebrovascular event were reviewed in conference by the investigators of this study in the absence of source data regarding the echocardiographic study or other investigations such as 24-h ambulatory BP. Stroke and transient ischemic attack (TIA) were prespecified end points at the beginning of the PIUMA study. Stroke was defined as a new neurologic deficit lasting at least 24 h, in the absence of underlying potentially important nonvascular causes. Patients with stroke were hospitalized during the acute phase. Brain imaging and other diagnostic tests were carried out for determination of stroke type. A TIA was diagnosed by a neurologist or internist in the presence of a rapid onset of a focal neurologic deficit lasting between 30 sec and 24 h, and presumably due to ischemia. The PIUMA protocol required that the deficit had to last until the time of the qualifying clinical examination to be accepted and coded as a terminating event. Data Analysis Statistical analysis was performed using SPSS (SPSS Inc., Chicago, IL) and SAS-Stat (SAS Institute, Cary, NC). Parametric data are reported as mean standard deviation. A paired t test was used to compare continuous variables between baseline and follow-up. Categorical variables were compared by a paired 2 test. Antihypertensive drugs were sorted into nonmutually exclusive categories and subjects receiving multidrug therapy were assigned to multiple categories according to the component drugs. For survival analyses, event-free curves were estimated using Kaplan-Meier product-limit method and compared by the Mantel (log-rank) test. For the subjects who experienced multiple events, survival analysis was based on the first event. The independent effect of several prognostic factors on survival was tested by stepwise Cox model. 15 The BP entered the model as either office BP or average 24-h ambulatory BP and the measure that yielded the best improvement of the 2 likelihood ratio was retained. Other tested variables were the serial changes in the ECG and echocardiographic features of LVH (discussed later), incident atrial fibrillation during follow-up,

3 AJH May 2006 VOL. 19, NO. 5 REGRESSION OF HYPERTROPHY AND STROKE 495 Table 1. Clinical characteristics at entry and follow-up Baseline visit Follow-up visit P Office systolic BP (mm Hg) 155 (18) 144 (17).001 Office diastolic BP (mm Hg) 98 (9) 90 (10).001 Office heart rate (beats/min) 75 (10) 73 (11) h systolic BP (mm Hg) 137 (14) 128 (12) h diastolic BP (mm Hg) 87 (10) 81 (8) h heart rate (beats/min) 75 (9) 73 (9).001 Daytime systolic BP (mm Hg) 142 (14) 133 (12).001 Daytime diastolic BP (mm Hg) 92 (10) 86 (9).001 Daytime heart rate (beats/min) 79 (9) 77 (9).001 Nighttime systolic BP (mm Hg) 126 (15) 119 (13).001 Nighttime diastolic BP (mm Hg) 78 (11) 73 (12).001 Nighttime heart rate (beats/min) 68 (9) 66 (9).001 Body weight (kg) 75.5 (14) 76.4 (14).001 Body mass index (kg/m 2 ) 26.6 (4) 26.9 (6).005 Glucose (mmol/l) 5.52 (1.05) 5.71 (1.71).001 Creatinine ( mol/l) 86.6 (18) 88.2 (30).074 Total cholesterol (mmol/l) 5.53 (1.06) 5.69 (1.18).001 HDL cholesterol (mmol/l) 1.27 (0.31) 1.35 (0.53).001 LDL cholesterol (mmol/l) 3.55 (0.94) 3.58 (0.95).464 Triglycerides (mmol/l) 1.63 (1.09) 1.70 (1.17).022 Uric acid (mmol/l) (0.08) (0.08).009 Potassium (mmol/l) 4.13 (0.36) 4.09 (0.41).118 LV hypertrophy at ECG (%) Interventricular septum (cm) 1.13 (0.23) 1.05 (1.19).001 LV internal diameter (cm) 4.93 (0.51) 4.96 (0.53).085 Posterior wall (cm) 1.01 (0.18) 0.98 (0.16).001 LV mass (g) 200 (63) 187 (57).001 LV mass (g/height[m] 2.7 ) 48.8 (14) 45.7 (12).001 LV hypertrophy at echocardiography (%) ECG electrocardiography; LV left ventricular. See text for details. Data expressed as mean (SD). age, gender (men, women), body mass index (kg/m 2 ), diabetes (no, yes) total cholesterol (mmol/l), serum triglycerides (mmol/l), family history of cardiovascular disease at age 55 years in the father or 65 years in the mother (no, yes), smoking habits (nonsmokers, current smokers), type of antihypertensive treatment at the follow-up visit, and statin treatment at the follow-up visit. Four groups defined by a normal LVM both at entry and follow-up (group A), increased LVM at entry and normal LVM at follow-up, denoting LVH regression (group B), increased LVM both at entry and follow-up, denoting lack of LVH regression (group C), and normal LVM at entry and increased LVM at follow-up, denoting new development of LVH (group D). A similar classification was used to define the serial changes in the ECG features of LVH. In two-tailed tests, P values.05 were considered statistically significant. Results We examined 880 subjects (43% women). Their mean age at entry was 48.2 years (SD 11 years). Prevalence of type 2 diabetes was 4.3% and that of current smokers was 25.5%. At entry, 71.1% of subjects had never been treated with antihypertensive drugs, whereas the remaining subjects had discontinued previous treatments from at least 4 weeks. Mean weight was 75.5 kg (SD 13 kg) and body mass index was 26.6 kg/m 2 (SD 3.7 kg/m 2 ). As shown in Table 1, prevalence of LVH was 14.4% by ECG and 36.6% by echocardiography. White coat hypertension (WCH), defined by a daytime BP 130 mm Hg systolic and 80 mm Hg diastolic, was found in 4.1% of the subjects (n 36). The LVM was 40.6 g/height 2.7 in the subset with WCH and 49.2 g/height 2.7 in the subset with ambulatory hypertension (P.001). Prevalence of LVH in the two groups was 16.1% and 37.4%, respectively (P.016). Changes During Follow-Up Table 1 shows the changes in the main clinical characteristics from the baseline to the follow-up visit, which was performed after a median of 3.5 years, still in the absence of cardiovascular events. At the follow-up visit, 29.6% of subjects were being treated with lifestyle measure only, whereas the remaining 70.4% were receiving antihypertensive drugs. After sorting antihypertensive treatments into nonmutually exclusive categories, frequency of subjects treated with thiazide diuretics, -blockers, ACE inhibitors, AT-II antagonists, calcium-channel blockers, and

4 496 REGRESSION OF HYPERTROPHY AND STROKE AJH May 2006 VOL. 19, NO blockers was 31.9%, 17.0%, 36.3%, 7.5%, 22.5%, and 3.9%., respectively. Treatment included only traditional antihypertensive drugs (thiazide diuretics and -blockers alone or combined) in 11.5% of subjects. At the follow-up visit, only 7.2% of subjects were being treated with a statin. As shown in Table 1, there was a consistent and statistically significant reduction in office and 24-h ambulatory BP (all P.001) and a slight increase in body weight (P.001) and body mass index (P.005). Among routine biochemical parameters, there was a slight increase in glucose (P.001) and uric acid (P.009) levels. The increase in total cholesterol (P.001) was accounted for by a consistent increase in the HDL fraction (P.001), whereas the LDL fraction did not change significantly (P.464). Prevalence of LVH diagnosed by ECG decreased from 14.4% to 11.4% (P.001). The reduction in LVM by echocardiography (P.001) was accounted for by a reduction in the thickness of both septum (P.001) and posterior wall (P.001), whereas the LV internal diameter did not change significantly. In univariate analyses, the changes in LVM corrected by height 2.7 correlated more closely (P.01) with the changes in 24-h systolic BP (r 0.39) than with those in office systolic BP (r 0.32). Cerebrovascular Complications An average of 2.4 years after the follow-up visit, 34 subjects experiences a first cerebrovascular event (stroke in 21, TIA in 13). Stroke was hemorrhagic in 2 subjects. In 33 of these subjects, stroke or TIA was the first cardiovascular event occurring during follow-up. An additional patient experienced a nonfatal myocardial infarction about 4 months before the stroke. Total duration of follow-up was 7.2 years (range, 0 to 18 years). Fig. 1 shows the event-free survival curves and the crude rate of events in four groups defined by the serial changes in the echocardiographic features of LVH (see Data Analysis section). The log-rank test was highly significant for trend (P.0003), FIG. 1. Risk of cerebrovascular events in relation to the previous changes in the echocardiographic features of left ventricular hypertrophy. FIG. 2. Risk of cerebrovascular events in relation to the previous changes in the electrocardiographic (left) or echocardiographic (right) features of left ventricular hypertrophy. See text for details regarding definition of left ventricular hypertrophy. but no difference was detected between groups A and B (P.65), as well as between groups C and D (P.41). Thus, group A was merged with B, and group C with D to combine regression of LVH and persistently normal LVM, as opposed to lack of regression of LVH and its new development. Fig. 2 shows the marked difference in the risk of cerebrovascular events between the two groups, with a crude rate of events of 0.25 events 100 subjects per year in the former and 1.16 events 100 subjects per year in the latter group (log-rank test: P.00001). Fig. 2 also shows that the serial changes in LVH diagnosed by ECG were unable to define groups at different risk of events (0.42 v 0.79 events 100 subjects per year; P.250), despite a favorable trend in the former group. The two curves crossed at the end of the follow up period, where only a minority of subjects remained into the study. None of the 34 subjects with WCH experienced a TIA or stroke during follow-up. Multivariate Analysis In a multivariate Cox analysis (Table 2), the risk of cerebrovascular events was 2.8 times higher (95% CI: ) in the subset with lack of regression or new development of LVH than in that with LVH regression or persistently normal LVM. Such effect was independent of age (P.001) and 24-h systolic BP at the follow-up visit (P.004). Office systolic BP at entry (P.15) and follow-up (P.09) and the average 24-h systolic BP at entry (P.075) did not achieve significance. None of the other tested variables (see Data Analysis section) achieved statistical significance. The relative risk independently associated with 1 SD increase in LVM at entry was 1.24 (95% CI ; P.26). Even the changes in LVM from baseline to follow-up and those in the Cornell voltage from baseline to follow-up did not achieve significance. None of the 18 subjects who experienced one or more episodes of atrial fibrillation showed a subsequent cerebrovascular event. The 5-year age-adjusted probability of cerebrovascular events in relation to the average 24-h systolic BP and LVM at the follow-up visit is shown in Fig. 3.

5 AJH May 2006 VOL. 19, NO. 5 REGRESSION OF HYPERTROPHY AND STROKE 497 Table 2. Independent predictors of cerebrovascular events Variable Comparison RR (95% CI) P Age 1 SD (10 y) 2.36 ( ) h systolic BP at follow-up 1 SD (12 mm Hg) 1.68 ( ).004 Serial changes in LVH 1 groups C D 2.80 ( ) groups A B BP blood pressure; groups A B the subjects with persistently normal left ventricular mass or regression of LVH; groups C D the subjects with lack of regression or new development of LVH; RR relative risk. Other tested variables (see data analyses) did not yield significance to enter the model. Discussion This study demonstrates that serial changes in LVH parallel the incidence of a composite pool of cerebrovascular events including stroke and TIA in a population of relatively young and initially untreated subjects with essential hypertension with and without LVH at entry. Even after adjustment for age and, most important, 24-h ambulatory BP, the link between regression of echocardiographic features of LVH and risk for cerebrovascular disease remained strong and independent. Previous Studies of LVH Regression Despite the diffusion of quantitative echocardiography, only a few longitudinal studies gathered a sample size large enough to investigate the potential prognostic impact of serial changes in the ECG and echocardiographic features of LVH in hypertensive subjects. This relative paucity is due to the low rate of major cardiovascular events in hypertensive subjects (approximately 2 events per 100 person-years) and, perhaps more important, to the need of collecting many readable echocardiographic tracings obtained before and during treatment, still before the occurrence of events. FIG. 3. Five-year age-adjusted probability of cerebrovascular events in relation to the average 24-h systolic blood pressure and left ventricular mass at the follow-up visit. Thus, although more than two decades have elapsed from the first description of LVH regression during antihypertensive treatment, 16 only a few studies were adequate in size and design to test the prognostic value of serial electrocardiographic 7 and echocardiographic 8,12 changes. Because the outcome measure in these studies was a composite pool of cardiovascular events including cardiac, cerebrovascular, and peripheral vascular complications, 8,12 the statistical power of these studies was generally not adequate to ascertain whether the overall benefit of LVH regression applied to the different sites of organ damage. On the other hand, LVH regression exerts favorable effects on coronary circulation, arrhythmias, and cardiac function, 17,18 whereas the potential basic mechanisms underlying the protection from the risk of stroke remain elusive. Quite recently two prespecified analyses of the LIFE study showed that the patients with lower levels of ECG 19 or echocardiographic 20 indices of LVH were at lesser risk of a composite end point of cardiovascular death, myocardial infarction, and stroke than those with higher levels. In these analyses, the prognostic effect of the serial changes in the ECG and echocardiographic indices was independent of several confounders including BP and assigned treatment. 19,20 LVH and Stroke In hypertensive subjects, an association was noted between LVM and carotid atherosclerosis. 3 A link between LVH and the risk of stroke has been shown in longitudinal studies of elderly subjects at increased risk of cerebrovascular events. 4,5 In a previous study from our group, initially untreated subjects with essential hypertension were followed for a mean of 5 years and LVH diagnosed at entry by ECG or echocardiography was associated with an excess risk for stroke and TIA, independently from office and ambulatory BP and other individual risk factors including atrial fibrillation. 6 However, all these studies left unanswered the important question of whether LVH regression per se confers a lesser risk for stroke. In the aforementioned analyses of the LIFE study, 19,20 the large sample size allowed separate assessment of the prognostic impact of LVH regression on cardiac and cerebrovascular complications. For each 1050 mm msec reduction in the Cornell product and 10.5 mm reduction in the Sokolow-Lyon voltage, the risk of stroke decreased by

6 498 REGRESSION OF HYPERTROPHY AND STROKE AJH May 2006 VOL. 19, NO. 5 10% (P.002) and 19% (P.001), respectively, after adjusting for treatment effect, baseline Framingham risk score, and systolic and diastolic BP at baseline and during treatment. 19 Similarly, for each 25.3 g (normalized by body surface area in square meters) reduction in in-treatment LVM, the risk of stroke decreased by 24% (P.02) after adjustment for assigned treatment, BP reduction, and baseline LVM. 20 The Present Study The present study extends the conclusion of the LIFE study to a large and relatively unselected population of hypertensive subjects with and without LVH (ie, patients who are commonly encountered in the clinical practice). The incident cerebrovascular events that occurred during the subsequent follow-up period were related to the prior changes in the ECG and echocardiographic features of LVH. A further strength of this study was the availability of 24-h ambulatory BP at baseline and follow-up in all subjects, which allowed us to improve the estimate of the predictive value of BP, thereby allowing a more accurate assessment of the independent prognostic value of serial changes in the ECG and echocardiographic features of LVH. The changes in LVM during treatment correlated more closely with the concomitant changes in ambulatory BP than with the changes in office BP, in line with previous reports. 21 Ambulatory BP better reflects the burden of BP load to which the left ventricle is exposed during daily life 21,22 and is more potent than office BP for prediction of cardiovascular disease. 23,24 In the present study, none of the subjects with WCH experienced a cerebrovascular events during follow-up. However, the present study was not designed, and thus was not adequately powered, to investigate the association between WCH and cerebrovascular events. In a large international database, which includes also the PIUMA study, WCH was not associated with a definitely increased risk of stroke, although a trend toward an increasing risk of cerebrovascular events was noted in the long run, by the ninth year of follow-up. 25 Mechanisms of Benefit The potential mechanisms underlying the prognostic impact of serial changes in LVM on the subsequent risk of cerebrovascular events remain speculative. In stroke-prone spontaneously hypertensive rats, inhibition of the reninangiotensin-system by ACE inhibitors or AT-II antagonists was associated with prevention of LVH and an increase in the life span, 26 which matched that of spontaneously normotensive rats. 26 Atrial fibrillation, which is predicted by LVH in hypertensive subjects, 27 was unlikely to mediate the link between LVH regression and the risk of stroke in our study because all of the 18 subjects who experienced one or more episodes of atrial fibrillation during follow-up remained free of subsequent cerebrovascular events. The impact of BP as a major determinant of the risk of stroke was controlled for in the multivariate analysis. However, it could be argued that even 24-h ambulatory BP might be inadequate to reflect the burden of BP fluctuations in the very long term. Alternatively, because of the poor association between BP and LVM, 12,21,22 LVH regression might reflect a reduced level of activity of a variety of factors that, in addition to BP, are known to affect both the growth of LVM and the progression of atherosclerosis. 1 Role of ECG Echocardiography is more sensitive than ECG for diagnosis of LVH. 13,28 In the present study, the serial changes in the ECG features of LVH were unable to identify groups at different risk of stroke during the entire follow-up period. As shown in Fig. 1, survival curves crossed at the end of follow-up, when only a minority of subjects remained into study. In a previous study from our group, the baseline ECG features of LVH predicted the subsequent incidence of stroke 6 and evidence is accumulating that the serial changes in the ECG and echocardiographic features of LVH are correlated. 29 Thus, it is likely that the present study was underpowered to test for the impact of serial ECG changes on the subsequent risk of stroke. Such analysis was possible in the larger setting of the LIFE study. 30 Limitations of the Study The total number of cerebrovascular events was relatively small. However, subjects with lack of regression or new development of echocardiographic LVH showed a severalfold increase in the risk of cerebrovascular events when compared to the subjects with absence of LVH or its regression. Given the high degree of statistical significance, it is unlikely that such an effect is due to chance. Furthermore, such an effect remained significant in the multivariate analysis with an adjusted excess risk of 2.8 times. The lack of fatal strokes may be accounted for by the relatively young age of the subjects, the relatively well-controlled BP during follow-up, the low prevalence of concomitant risk factors, and the absence of previous cardiovascular morbid events. The TIAs are known to be relatively soft end points. However, because TIAs had to be symptomatic at the time of diagnosis by a neurologist or internist, their incidence may have been underestimated. Finally, because the PIUMA study is conducted in a sample entirely composed by white subjects, caution is needed in extrapolating results to different ethnic groups. Implications Our results strongly suggest that LVH regression is a surrogate marker of decreased risk for cerebrovascular complications in hypertensive patients. Although limited by the relatively small sample size, this study suggests the clinical usefulness of periodic measurements of LVM in patients with essential hypertension. The main clinical message is that patients who fail to achieve LVH regression or develop new LVH should be considered at

7 AJH May 2006 VOL. 19, NO. 5 REGRESSION OF HYPERTROPHY AND STROKE 499 increased risk for subsequent cerebrovascular disease. Future trials should investigate the potential benefit of a more aggressive therapeutic strategy in these patients. Acknowledgment We gratefully thank Francesca Saveri for secretarial assistance. References 1. Benjamin EJ, Levy D: Why is left ventricular hypertrophy so predictive of morbidity and mortality? Am J Med Sci 1999;317: Bahemuka M: Primary intracerebral hemorrage and heart weight: a clinicopathologic case-control review of 218 patients. Stroke 1987; 18: Okin PM, Roman MJ, Devereux RB, Kligfield P: Association of carotid atherosclerosis with electrocardiographic myocardial ischemia and left ventricular hypertrophy. Hypertension 1996;28: Bikkina M, Levy D, Evans JC, Larson MG, Benjamin EJ, Wolf PA, Castelli WP: Left ventricular mass and risk of stroke in an elderly cohort. The Framingham Heart Study. JAMA 1994;272: Aronow WS, Ahn C, Kronzon I, Gutstein H, Schoenfeld MR: Association of extracranial carotid arterial disease, prior atherothrombotic brain infarction, systemic hypertension, and left ventricular hypertrophy with the incidence of new atherothrombotic brain infarction at 45-month follow-up in 1,482 older patients. Am J Cardiol 1997;79: Verdecchia P, Porcellati C, Reboldi G, Gattobigio R, Borgioni C, Pearson TA, Ambrosio G: Left ventricular hypertrophy as an independent predictor of acute cerebrovascular events in essential hypertension. Circulation 2001;104: Levy D, Salomon M, D Agostino RB, Belanger AJ, Kannel WB: Prognostic implications of baseline electrocardiographic features and their serial changes in subjects with left ventricular hypertrophy. Circulation 1994;90: Verdecchia P, Angeli F, Borgioni C, Gattobigio R, de Simone G, Devereux RB, Porcellati C: Changes in cardiovascular risk by reduction of left ventricular mass in hypertension: a meta-analysis. Am J Hypertens 2003;16: Okin PM, Devereux RB, Jern S, Kjeldsen SE, Julius S, Nieminen MS, Snapinn S, Harris KE, Aurup P, Edelman JM, Dahlof B, for the Losartan Intervention for Endpoint reduction in hypertension Study Investigators: Regression of electrocardiographic left ventricular hypertrophy by losartan versus atenolol: The Losartan Intervention for Endpoint reduction in Hypertension (LIFE) study. Circulation 2003;108: Dahlof B, Devereux RB, Kjeldsen SE, Julius S, Beevers G, de Faire U, Fyhrquist F, Ibsen H, Kristiansson K, Lederballe-Pedersen O, Lindholm LH, Nieminen MS, Omvik P, Oparil S, Wedel H, for the LIFE Study Group: Cardiovascular morbidity and mortality in the Losartan Intervention For Endpoint reduction in hypertension study (LIFE): a randomised trial against atenolol. Lancet 2002;359: Verdecchia P, Angeli F, Reboldi G, Carluccio E, Benemio G, Gattobigio R, Borgioni C, Bentivoglio M, Porcellati C, Ambrosio G: Improved cardiovascular risk stratification by a simple ECG index in hypertension. Am J Hypertens 2003;16: Verdecchia P, Schillaci G, Borgioni C, Ciucci A, Gattobigio R, Zampi I, Reboldi G, Porcellati C: Prognostic significance of serial changes in left ventricular mass in essential hypertension. Circulation 1998;97: Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E, Sachs I, Reichek N: Echocardiographic assessment of left ventricular hypertrophy: comparison to necropsy findings. Am J Cardiol 1986; 57: de Simone G, Daniels SR, Devereux RB, Meyer RA, Roman MJ, de Divitiis O, Alderman MH: Left ventricular mass and body size in normotensive children and adults: assessment of allometric relations and impact of overweight. J Am Coll Cardiol 1992;20: Cox DR: Regression models and life-tables. J R Stat Soc (B) 1972;34: Fouad FM, Nakashira P, Tarazi RS, Salcedo EE: Reversal of left ventricular hypertrophy in hypertensive patients treated with methyldopa. Am J Cardiol 1982;49: Motz W, Strauer BE: Therapy of hypertensive cardiac hypertrophy and impaired coronary microcirculation. J Cardiovasc Pharmacol 1994;24(Suppl 1):S34 S Wachtell K, Palmieri V, Olsen MH, Gerdts E, Papademetriou V, Nieminen MS, Smith G, Dahlof B, Aurigemma GP, Devereux RB: Change in systolic left ventricular performance after 3 years of antihypertensive treatment: the Losartan Intervention for Endpoint (LIFE) study. Circulation 2002;106: Okin PM, Devereux RB, Jern S, Kjeldsen SE, Julius S, Nieminen MS, Snapinn S, Harris KE, Aurup P, Edelman JM, Wedel H, Lindholm LH, Dahlof B, for the LIFE Study Investigators: Regression of electrocardiographic left ventricular hypertrophy during antihypertensive treatment and the prediction of major cardiovascular events. JAMA 2004;106: Devereux RB, Wachtell K, Gerdts E, Boman K, Nieminen MS, Papademetriou V, Rokkedal J, Harris K, Aurup P, Dahlof B: Prognostic significance of left ventricular mass change during treatment of hypertension. JAMA 2004;292: Verdecchia P, Reboldi G, Porcellati C, Schillaci G, Pede S, Bentivoglio M, Angeli F, Norgiolini S, Ambrosio G: Risk of cardiovascular disease in relation to achieved office and ambulatory blood pressure control in treated hypertensive subjects. J Am Coll Cardiol 2002;39: Verdecchia P, Schillaci G, Guerrieri M, Gatteschi C, Benemio G, Boldrini F, Porcellati C: Circadian blood pressure changes and left ventricular hypertrophy in essential hypertension. Circulation 1990; 81: Verdecchia P: Prognostic value of ambulatory blood pressure: current evidence and clinical implications. Hypertension 2000;35: Clement DL, De Buyzere ML, De Bacquer DA, de Leeuw PW, Duprez DA, Fagard RH, Gheeraert PJ, Missault LH, Braun JJ, Six RO, Van Der Niepen P, O Brien E, for the Office versus Ambulatory Pressure Study Investigators: Prognostic value of ambulatory bloodpressure recordings in patients with treated hypertension. N Engl J Med 2003;348: Verdecchia P, Reboldi GP, Angeli F, Schillaci G, Schwartz JE, Pickering TG, Imai Y, Ohkubo T, Kario K: Short- and long-term incidence of stroke in white-coat hypertension. Hypertension 2005;45: Linz W, Jessen T, Becker RHA, Schölkens BA, Wiemer G: Longterm ACE inhibition doubles lifespan of hypertensive rats. Circulation 1997;96: Verdecchia P, Reboldi G, Gattobigio R, Bentivoglio M, Borgioni C, Angeli F, Carluccio E, Sardone MG, Porcellati C: Atrial fibrillation in hypertension: predictors and outcome. Hypertension 2003;41: Devereux RB, Casale PN, Wallerson DC, Kligfield P, Hammond IW, Liebson PR, Campo E, Alonso DR, Laragh JH: Cost-effectiveness of echocardiography and electrocardiography for detection of left ventricular hypertrophy in patients with systemic hypertension. Hypertension 1987;9:II69 II Devereux RB: Detection of left ventricular hypertrophy by M- mode echocardiography. Anatomic validation, standardization, and comparison to other methods. Hypertension 1987;9(Suppl II):II9 II Okin PM, Devereux RB, Liu JE, Oikarinen L, Jern S, Kjeldsen SE, Julius S, Wachtell K, Nieminen MS, Dahlof B: Regression of electrocardiographic left ventricular hypertrophy predicts regression of echocardiographic left ventricular mass: the LIFE study. J Hum Hypertens 2004;18: