Abstract
Left ventricular hypertrophy (LVH) is often used as an indicator to assess hypertension‐mediated organ damage (HMOD), alongside hypertensive retinopathy (HR) and nephropathy. Assessment of HMOD is crucial when making decisions about treatment optimization. Despite longstanding debate over its reliability to detect LVH, it is common practice to perform an electrocardiogram (ECG) instead of directly assessing left ventricular mass with echocardiography. In this study, the presence of LVH was evaluated using both ECG and echocardiography among consecutive patients suspected of therapy‐resistant hypertension or secondary hypertension in the outpatient clinic of the Department of Internal Medicine at the Diakonessen Hospital, Utrecht, the Netherlands, between July 15, 2017, and July 31, 2020. The primary endpoints were the specificity and sensitivity of ECG as a diagnostic tool for LVH, with echocardiography serving as the reference method. Among the 329 participants, we identified 70 individuals (21.3%) with true LVH based on echocardiography. The ECG displayed a sensitivity of 47.9% and a specificity of 75.3%. Moreover, the area under the receiver operating characteristics curve was 0.604. In conclusion, ECG demonstrates limited value in identifying LVH. Considering the importance of accurately assessing HMOD for treatment optimization of hypertension, the role of ECG as a diagnostic tool for LVH is, therefore, questionable. Instead, we recommend employing standard echocardiography as a more reliable diagnostic.
Keywords: clinical management of high blood pressure (HBP), left ventricular hypertrophy (LVH) hypertension/arrhythmias, resistant hypertension, secondary high blood pressure (HBP), treatment and diagnosis/guidelines
1. INTRODUCTION
Uncontrolled hypertension can lead to hypertension‐mediated organ damage (HMOD), which refers to structural or functional changes caused by elevated blood pressure in arteries or end organs such as the heart, blood vessels, brain, eyes, and kidneys. 1 Accurately monitoring HMOD is essential for treatment optimization and can be done through examination of a patient's medical history and screening for hypertensive retinopathy (HR), nephropathy and left ventricular hypertrophy (LVH). 1 , 2 , 3 LVH is furthermore an independent risk factor for cardiovascular (CV) events. 4 , 5 , 6 , 7 Early recognition of LVH and adequate antihypertensive treatment can induce regression of LVH and help avoid irreversible LVH. 8 , 9 , 10 , 11 , 12 Electrocardiogram (ECG) is often used for LVH screening since it is noninvasive, cheap, and readily available. However, it has been consistently reported that the sensitivity and specificity of ECG for diagnosing LVH are rather low. 13 , 14 , 15 Consequently, hypertensive patients are at risk of being misclassified for having LVH when relying solely on ECG. This study assesses the value of ECG in screening for LVH in patients suspected of therapy‐resistant hypertension or secondary hypertension, with echocardiography serving as the reference standard.
2. SUBJECTS AND METHODS
2.1. Population
A cross‐sectional study was used to obtain data from consecutive patients suspected of therapy‐resistant hypertension or secondary hypertension at the outpatient clinic of the Department of Internal Medicine of the Diakonessen Hospital, Utrecht, the Netherlands, between July 15, 2017, and July 31, 2020. All data were collected within a maximum of 2 weeks for each patient. For the vast majority of patients, data collection took place on 1 day. Patients whose ECG and echocardiography were obtained more than 2 weeks before or after the day of data collection were not included. Additionally, patients for whom ECG or echocardiography data were not available or of insufficient quality were excluded.
All patients who were included underwent assessment of cardiovascular (CV) risk factors, 24‐h ambulatory and office measurement of blood pressure, as well as drug history and presence of HMOD. Patient data were anonymized and de‐identified before analyses; as such, the need for written informed patient consent was waived by the institutional review board.
2.2. Baseline assessment
Blood pressure was measured three times on both arms at the physician's office in a well‐lit and warm room while the patient was sitting in a chair with the patient's arm placed on a table using a fully automated upper arm blood pressure monitor, operating on the oscillometric principle (Omron, Omron Healthcare Europe Inc., Hoofddorp, The Netherlands). Only the final measurement of the arm with the highest blood pressure was used. In all, 24‐h ambulatory blood pressure monitors operating on the oscillometric principle determined blood pressure three times per hour between 07:00 and 23:00 h, and once per hour between 23:00 and 07:00. Mean systolic blood pressure and diastolic blood pressure were calculated using specialized computer software (Spacelabs ABP, Ite Medical, Tiel, The Netherlands). Hypertension was defined according to international guidelines as the current use of antihypertensive medication and/or blood pressure above 135/85 mmHg for the ambulatory measurement during the day, blood pressure above 120/70 mmHg for the measurement during the night or blood pressure above 130/80 mmHg for the total measurement in 24 h. 1 Therapy‐resistant hypertension was defined as a blood pressure that remains above goal, despite the use of at least three antihypertensive drugs.
Information about prior cardiac disease was based on the participant's medical records, obtained from the participant's anamnesis, the general practitioner and hospital history. Prior cardiac disease included myocardial infarction, angina pectoris, congestive heart failure, cardiac arrythmias and valve insufficiency.
2.3. Assessment of HMOD
Retinal fundoscopy was performed to evaluate HR. Thirty minutes before fundoscopy, pupils were dilated. Indirect fundoscopy was performed by a single ophthalmologist for each patient. HR was classified into four grades according to the system previously described by Keith and colleagues. 16
To determine presence of microalbuminuria, each patient collected 24‐h urine. Albumin was measured in a 24‐h urine sample. Microalbuminuria was defined as albumin ≥30 mg/24 h. 1
In each patient a 12‐lead ECG was obtained to evaluate presence of LVH. Among the ECG criteria used are [1] the Sokolow–Lyon criteria, [2] the voltage criteria according to Cornell, and [3] the Cornell product. 17 , 18 Furthermore, the following [4] conventional voltage criteria were used: R > 25 mm in V5 or V6, R > 20 mm in II or III, R > 14 mm in I, R > 12 mm in aVL, on the condition of the absence of a left bundle branch block. Next to classic voltage criteria, also [5] nonvoltage criteria for left ventricular strain pattern were used, defined as ST‐segment depression and T‐top‐inversion. Additionally, conditions like hypertrophic cardiomyopathy, valvular heart disease and left or right bundle branch block were assessed, although they were not utilized as exclusion criteria. The ECGs were interpreted by a single cardiologist for each patient during the outpatient clinic visit and a second time by the study team.
2.4. Echocardiography
A transthoracic echocardiogram was made using two‐dimensional images and M‐mode. Echocardiography was performed by a technician and reviewed by a cardiologist. As a first step, left ventricular (LV) diameter, septal wall thickness and posterior wall thickness were measured in parasternal long‐axis view. To assess presence of LVH, the left ventricular mass (LVM) was measured using the ASE/EACVI formula: LVM = 0.8 × [1.04((LVIDd + IVSd + PWT)3– LVIDd3)] + 0.6, with LVIDd indicating left ventricle internal dimension diastole, IVSd; interventricular septum diastole and PW; posterior wall thickness. 19 The outcome was corrected for the Body Surface Area (BSA). A normal LVM for men was defined as ≤115 g/m2 and for women ≤ 95 g/m2. The relative wall thickness (RWT) was calculated by the ratio of twice the PW and the LV diastolic diameter. Concentric LVH was defined as LVM above normal values and a RWT > 0.42.
Additional structures of the heart were visualized to obtain information about the heart's functioning. The ejection fraction was measured using Simpson's biplane method and/or utilizing a 3D heart model. Ventricular systolic function was interpreted by the cardiologist as normal or reduced. Furthermore, insufficiency or stenosis of the mitral, tricuspid, pulmonary, and aortic valves were evaluated using doppler technique. The severity of the insufficiency was scored as no, minor, moderate or severe. Diastolic function was evaluated using the recommended ASA/EACVI standard which includes the ratio of peak velocity of early mitral inflow to the peak velocity of late mitral inflow (E/A ratio), early diastolic mitral inflow velocity to early diastolic mitral annulus velocity (E/e' ratio), tricuspid regurgitation velocity and indexed left atrial volume. 20
2.5. Statistical analyses
Characteristics of the study population are presented as mean ± standard deviation (SD) or percentages in the case of categorical data. Student's t ‐test or Mann–Whitney U‐test were used to compare numerical data. Pearson's Chi‐squares test followed by a pairwise Z‐test was used to compare categorical data. All analyses were carried out using SPSS 26.0 software (SPSS Inc., Chicago, IL) and p < .05 were considered as statistically significant.
3. RESULTS
In total 335 patients were included, of which six had to be excluded from the analyses based upon insufficient quality of echocardiography images. Patient characteristics are shown in Table 1. The population had a mean age of 53.4 (±15.5) years and had an average BMI of 27.8 (±20.4) kg/m2. 44.4% of the patients were of male sex. 8.5% patients had a history of cardiac disease: 4 out of 329 patients previously had a myocardial infarction, eight were known with angina pectoris, eight with congestive heart failure, seven with cardiac arrythmias, and seven patients were known to a valve insufficiency.
TABLE 1.
| Basic characteristics of the patient population.
| Total (n = 329) | No LVH (n = 259) | LVH (n = 70) | p‐value | |
|---|---|---|---|---|
| Demographic factors | ||||
| Age (years) | 53.3 (±15.5) | 51.4 (±15.6) | 60.3 (±13.2) | <.001 |
| % male sex | 44.4% | 43.6% | 47.1% | .600 |
| Weight (kg) | 84.2 (±20.4) | 83.6 (±20.1) | 86.5 (±21.4) | .509 |
| BMI (kg/m2) | 27.8 (±6.2) | 27.6 (±6.1) | 28.6 (±6.7) | .163 |
| % prior cardiac disease | 8.5% | 7.3% | 12.9% | .142 |
| % smoking | 18.0% | 16.7% | 22.9% | .232 |
| % with diabetes | 12.5% | 11.2% | 17.1% | .181 |
| % using statins | 21.9% | 20.1% | 28.6% | .127 |
| Blood pressure factors | ||||
| Systolic BP (office) (mmHg) | 160.2 (±24.5) | 158.2 (±24.7) | 167.4 (22.5) | .004 |
| Diastolic BP (office) (mmHg) | 94.6 (±14.0) | 94.5 (±14.2) | 94.8 (±13.1) | .986 |
| Systolic BP (day) (mmHg) | 149.7 (±17.9) | 148.5 (±18.3) | 154.4 (±15.5) | .002 |
| Diastolic BP (day) (mmHg) | 89.4 (±12.6) | 89.3 (±12.8) | 89.4 (±12.5) | .866 |
| Systolic BP (night) (mmHg) | 133.0 (±20.8) | 131.3 (±19.8) | 139.7 (±23.1) | <.001 |
| Diastolic BP (night) (mmHg) | 76.8 (±12.1) | 79.1 (±13.5) | 76.2 (±11.7) | .116 |
| Systolic BP (total) (mmHg) | 144.6 (±17.1) | 143.0 (±17.4) | 150.9 (±14.6) | <.001 |
| Diastolic BP (total) (mmHg) | 85.5 (±11.9) | 85.3 (±11.7) | 86.1 (±12.5) | .676 |
| Number of antihypertensive drugs | .063 | |||
| 0 | 23.4% | 22.8% | 25.7% | n/a * |
| 1–2 | 40.4% | 43.6% | 28.6% | n/a * |
| ≥3 | 36.2% | 33.6% | 45.7% | n/a * |
| % therapy‐resistant | 21.3% | 18.5% | 31.4% | .019 |
| Non‐cardiac markers of end‐organ damage | ||||
| % with microalbuminuria | 24.6% | 22.4% | 32.9% | .071 |
| % with hypertensive retinopathy | 30.1% | 29.1% | 33.3% | .503 |
| Biochemic factors | ||||
| Glucose (mmol/L) | 6.00 (±1.42) | 5.96 (±1.48) | 6.13 (±1.17) | .016 |
| HbA1c (mmol/mol) | 38.1 (±8.4) | 38.0 (±9.0) | 38.3 (±5.7) | .058 |
| Total cholesterol (mmol/L) | 4.80 (±0.99) | 4.76 (±0.97) | 4.92 (±1.03) | .188 |
| LDL‐c (mmol/L) | 2.69 (±1.11) | 2.64 (±0.89) | 2.89 (±1.71) | .266 |
| HDL‐c (mmol/L) | 1.41 (±0.50) | 1.44 (±0.52) | 1.32 (±0.40) | .211 |
| Triglycerides (mmol/L) | 1.65 (±1.25) | 1.58 (+1.26) | 1.88 (±1.18) | <.001 |
| eGFR (mL/min/1.73 m2) | 84.3 (±18.1) | 86.1 (±17.5) | 77.6 (±19.0) | .003 |
| Positive ECG criteria for LVH | <.001 | |||
| 0 | 70.5% | 75.3% | 52.9% | <.001 |
| 1 | 16.4% | 16.2% | 17.1% | .853 |
| 2 | 8.2% | 5.0% | 20.0% | <.001 |
| 3 | 3.6% | 2.3% | 8.6% | .013 |
| 4 | 1.2% | 1.2% | 1.4% | .855 |
Notes: Based on transthoracic echocardiography a group with LVH and without LVH was separated. Values are presented as mean (±SD) or percentage. GFR was estimated using the Chronic Kidney Disease Epidemiology Collaboration equation. n/a indicates not applicable.
Abbreviations: BMI, body mass index; BP, blood pressure; eGFR; estimated glomerular filtration rate; LDL‐c, low‐density lipoprotein cholesterol; LVH, left ventricular hypertrophy; HDL‐c high‐density lipoprotein cholesterol.
A pairwise Z‐test was not performed in case Pearson's Chi‐squares test yielded a p > .05.
3.1. Echocardiography and LVH
Echocardiography showed concentric LVH in 70 out of the 329 patients (21.3%). As shown in Table 1, patients with LVH were older than patients without LVH, whereas there were no differences in sex or BMI. Furthermore, patients with LVH had a higher systolic blood pressure during the office, the night and the 24‐h measurement, and were more likely to be therapy‐resistant. The diastolic blood pressure did not differ. Albeit lacking statistical significance, the data suggest that individuals with LVH are more likely to also have HR and microalbuminuria. Additionally, it was observed that patients with LVH exhibited impaired kidney function as determined by the CKD‐EPI measurement. Moreover, individuals with LVH demonstrated elevated levels of glucose and triglycerides in their blood samples.
3.2. ECG and LVH
Based upon any ECG criteria LVH was diagnosed in 97 patients (29.5%). Of these patients, 54 had only one positive criterion for LVH, while 27, 12, and four patients had two, three, or four positive criteria, respectively. In Table 1, it is observed that patients without LVH, as confirmed by echocardiography, were more likely to have no positive ECG criteria for LVH. Additionally, patients with established LVH were more likely to exhibit 2 or 3 positive ECG criteria than those without LVH. Notably, the occurrence of a single positive ECG criterion did not differ between the two groups. Conventional voltage criteria were positive in 48 patients, with Sokolow–Lyon voltage criterion being positive in 35 patients, Cornell voltage criterion in 34 patients, Cornell duration product in 31 patients, and a strain pattern was present in 12 patients.
3.3. Discrepancy and congruency between ECG and echocardiography
The performance metrics of the ECG criteria for diagnosing LVH are shown in Table 2. When combining all ECG criteria, 33 of the 70 patients with LVH were detected by ECG, which translates to a sensitivity of 47.1%. In other words, half of the patients with LVH are missed in case ECG is used to rule out LVH (37/70). Of the 232 patients with negative criteria for LVH on the ECG, 195 patients truly had no LVH, meaning a specificity of 75.3%. One in every six patients with a normal ECG was misclassified and the presence of LVH was missed (37/232). In case ECG is used to diagnose LVH, two out of three are misdiagnosed and have no LVH (64/97). Furthermore, Table 2 delineates the specific characteristics of the individual ECG criteria for diagnosing LVH. The conventional criteria showed the highest sensitivity at 27.1%, whereas the strain pattern exhibited the lowest sensitivity at 7.2%. Conversely, the strain pattern had the highest specificity at 97.3%, whereas the conventional criteria had the lowest specificity at 88.8%.
TABLE 2.
| Performance metrics of various ECG criteria for diagnosing LVH.
| All ECG criteria | Sokolow–Lyon criteria | Cornell voltage criteria | Cornell product | Conventional criteria | Strain pattern | |
|---|---|---|---|---|---|---|
| Sensitivity | 0.471 | 0.143 | 0.214 | 0.186 | 0.271 | 0.072 |
| Specificity | 0.753 | 0.903 | 0.927 | 0.931 | 0.888 | 0,973 |
| PPV | 0.340 | 0.286 | 0.441 | 0.419 | 0.396 | 0.417 |
| NPV | 0.841 | 0.796 | 0.814 | 0.809 | 0.819 | 0.798 |
| AUC | 0.604 | 0.517 | 0.572 | 0.559 | 0.574 | 0.523 |
Notes: “All ECG criteria” denotes the diagnosis of LVH based upon any ECG criterium being positive.
Abbreviations: NPV, negative predictive value; PPV, positive predictive value; AUC, area under the curve, derived from the ROC curve as shown in Figure 1.
To analyze the discriminative capacity of each individual ECG criterion, as well as of all criteria combined, a receiver operator characteristic (ROC) analysis was conducted. The resulting ROC curves are illustrated in Figure 1. The area under the curve (AUC) for each criterion was calculated and is presented in Table 2. When all ECG criteria were combined, the AUC was 0.604. Among the individual criteria, the Cornell voltage criteria achieved the highest AUC at 0.572, whereas the Sokolow–Lyon criteria had the lowest AUC at 0.517.
FIGURE 1.

Receiver operating characteristics (ROC) curves. The ROC curves of the individual ECG criteria, as well as, of all ECG criteria combined are depicted. The AUCs derived from these curves are displayed in Table 2.
As presented in Table 3, it was found that patients who were incorrectly diagnosed with LVH based on ECG criteria, were relatively younger than those who were accurately diagnosed with LVH. In addition, these patients demonstrated lower levels of glucose and triglycerides, and better kidney function based on CKD‐EPI. Although the observed differences are not statistically significant, the data suggest that patients with true LVH may have a higher incidence of HR and microalbuminuria. No dissimilarities in the patient characteristics were identified that could effectively distinguish between the true positive and false negative groups.
TABLE 3.
| Characteristics of patients with a true positive and false positive ECG for LVH.
| True positive (n = 33) | False positive (n = 64) | p‐value | |
|---|---|---|---|
| Demographic factors | |||
| Age (years) | 58.9(±13.9) | 49.4 (±17.3) | .005 |
| % male | 51.5% | 56.3% | .657 |
| Weight (kg) | 84.8 (±20.9) | 83.6 (±21.0) | .841 |
| BMI (kg/m2) | 27.3 (±7.8) | 26.5 (±6.8) | .726 |
| % diabetes | 9.1% | 7.8% | .828 |
| % smoking | 24.2% | 23.8% | .962 |
| % statins | 30.3% | 23.4% | .464 |
| Blood pressure factors | |||
| Systolic BP (office) (mmHg) | 171.5 (±27.3) | 163.8 (±22.3) | .137 |
| Diastolic BP (office) (mmHg) | 95.9 (±11.4) | 96.8 ± 14.1) | .664 |
| Systolic BP (total) (mmHg) | 152.7 (±15.7) | 147.6 (±19.5) | .324 |
| Diastolic BP (total) (mmHg) | 87.3 (±14.4) | 87.8 (±12.2) | .637 |
| Number of antihypertensive drugs | .080 | ||
| 0 | 21.2% | 18.8% | n/a * |
| 1–2 | 24.2% | 46.9% | n/a * |
| ≥3 | 54.5% | 34.4% | n/a * |
| % therapy‐resistant | 30.3% | 15.6% | .090 |
| Non‐cardiac markers of end‐organ damage | |||
| % hypertensive retinopathy | 37.5% | 30.6% | .503 |
| % microalbuminuria | 45.5% | 31.3% | .168 |
| Biochemic factors | |||
| Glucose (mmol/L) | 6.1 (±0.9) | 5.7 (±0.9) | .017 |
| HbA1c (mmol/mol) | 38.4 (±4.8) | 36.9 (±5.5) | .048 |
| Total cholesterol (mmol/L) | 4.84 (±1.08) | 4.55 (±1.03) | .089 |
| LDL‐c (mmol/L) | 2.59 (±1.15) | 2.36 (±0.83) | .167 |
| HDL‐c (mmol/L) | 1.32 (±0.43) | 1.41 (±0.59) | .609 |
| Triglycerides (mmol/L) | 1.93 (±1.06) | 1.77 (±1.90) | .022 |
| eGFR (mL/min/1.73 m2) | 77.3 (±20.7) | 88.0 (±18.7) | .021 |
Notes: Values are presented as mean (±SD) or percentage. The true positive group consists of patients who meet the LVH criteria on both ECG and echocardiogram, while the false positive group includes patients with positive criteria for LVH on ECG but not on echocardiogram. GFR was estimated using the Chronic Kidney Disease Epidemiology Collaboration equation. n/a indicates not applicable.
Abbreviations: BMI, body mass index; BP, blood pressure; eGFR, estimated glomerular filtration rate; LDL‐c, low‐density lipoprotein cholesterol; LVH, left ventricular hypertrophy; HDL‐c high‐density lipoprotein cholesterol.
A pairwise Z‐test was not performed in case Pearson's Chi‐squares test yielded a p > .05.
When we compared patients in whom the ECG had missed the presence of LVH compared to the group that was correctly negative for LVH on ECG, the patients in the false negative group were significantly older and had a higher BMI (data not shown). Furthermore, these patients displayed a significantly higher 24‐h systolic blood pressure. No differences in other patient characteristics were identified.
3.4. Additional information on echocardiography
In eight patients left ventricular systolic dysfunction was found. Diastolic dysfunction was detected in 184 patients. Aortic valve stenosis was not detected in any of the patients. As shown in Table 4, patients with LVH were more likely to have LV diastolic dysfunction. Reduced ejection fraction below 50% was found in nine patients, of whom one was below 40%. No difference in ejection fraction between patients with and without LVH was observed.
TABLE 4.
| Additional findings from echocardiography when assessing LVH.
| Total (n = 329) | No LVH (n = 259) | LVH (n = 70) | p‐value | |
|---|---|---|---|---|
| Findings on echocardiogram | ||||
| Intraventricular septum thickness (cm) | 1.03 (±0.23) | 0.97 (±0.20) | 1.26 (±0.18) | <.001 |
| Left ventricular mass (gram/m2) | 88.4 (±27.4) | 80.7 (±18.9) | 116.7 (±34.8) | <.001 |
| % with impaired LV function | 2.5% | 2.0% | 4.3% | .255 |
| LV diastolic function | <.001 | |||
| Normal | 35.2% | 39.0% | 18.9% | .006 |
| Grade 1 dysfunction | 64.1% | 61.0% | 77.4% | .026 |
| Grade 2 dysfunction | 0.7% | 0.0% | 3.8% | n/a ** |
| Grade 3/4 dysfunction | 0.0% | 0.0% | 0.0% | n/a ** |
| Aortic valve insufficiency | .151 | |||
| % No | 89.8% | 91.3% | 84.1% | n/a * |
| % Minor | 9.9% | 8.3% | 15.9% | n/a * |
| % Moderate | 0.3% | 0.4% | 0.0% | n/a * |
| % Severe | 0.0% | 0.0% | 0.0% | n/a * |
| Pulmonary valve insufficiency | .616 | |||
| % No | 86.3% | 85.0% | 90.5% | n/a * |
| % Minor | 13.2% | 14.3% | 9.5% | n/a * |
| % Moderate | 0.5% | 0.7% | 0.0% | n/a * |
| % Severe | 0.0% | 0.0% | 0.0% | n/a * |
| Mitral valve insufficiency | .048 | |||
| % No | 63.8% | 66.7% | 53.0% | .041 |
| % Minor | 34.3% | 32.1% | 42.4% | .117 |
| % Moderate | 1.9% | 1.2% | 4.5% | .084 |
| % Severe | 0.0% | 0.0% | 0.0% | n/a ** |
| Triscupid valve insufficiency | .659 | |||
| % No | 71.8% | 71.3% | 73.8% | n/a * |
| % Minor | 27.2% | 27.5% | 26.2% | n/a * |
| % Moderate | 1.0% | 1.3% | 0.0% | n/a * |
| % Severe | 0.0% | 0.0% | 0.0% | n/a * |
| Ejection fraction (%) | 75.5 (±19.0) | 75.6 (±19.6) | 74.8 (±17.3) | .950 |
| % with concentric remodeling | 25.08% | 31.13% | 2.86% | <.001 |
Notes: Based on the presence of LVH on transthoracic echocardiography a group with LVH and without LVH was separated. Values are presented as mean (±SD) or percentage.
Abbreviations: LV, left ventricle; LVH, left ventricular hypertrophy; n/a, not applicable.
A pairwise Z‐test was not performed in case Pearson's Chi‐squares test yielded a p > .05.
Due to a proportion equaling zero or one, no pairwise Z‐test could be performed.
In addition, we detected regurgitation of the aortic, mitral, tricuspid, and pulmonary valves in 32, 112, 85, and 25 patients, respectively. As presented in Table 4, patients without LVH had a higher probability of not having mitral valve insufficiency compared to the patients with LVH. No other significant distinctions were found regarding the presence or absence of insufficiency of the aortic, tricuspid, or pulmonary valves.
3.5. Signs of HMOD in patients with hypertension
From all 329 patients, 182 showed one or more signs of HMOD. LVH displayed the lowest prevalence among the HMOD indicators at 21.3%, while HR exhibited the highest prevalence at 28.9%. 24.6% of the included patients exhibited microalbuminuria. It is noteworthy that a total of 37 patients (constituting 20.3% of all patients with HMOD) only showed LVH as an HMOD marker, thus would have been incorrectly classified as non‐HMOD if LVH detection was omitted.
4. DISCUSSION
In accordance with previous findings, the results presented in this article show that ECG has low specificity and especially low sensitivity for diagnosing LVH. 13 , 14 , 15 Furthermore, the AUCs for the individual ECG criteria, derived from the ROC curves, were proximate to 0.5, indicating that these criteria have a low capacity to discriminate between the presence and absence of LVH. Patients incorrectly diagnosed with LVH on the ECG showed some differences with those correctly diagnosed. However, no clear demarcated population was found in which ECG could be used reliably. We conclude that the ECG has very limited value in screening for LVH, as well as in determining the necessity for performing echocardiographic examination.
As demonstrated in this article, echocardiography is very suitable for the detection of LVH. When making the decision to initiate or intensify antihypertensive treatment, a detailed assessment of HMOD, including LVH, is required for optimal personalized management. This is particularly important in relatively younger patients, patients with high‐normal blood pressure and those who are therapy‐resistant. 1 , 21 Furthermore, accurate detection of LVH identifies a substantial group of patients in whom LVH is the sole marker of HMOD. Additionally, LVH serves as an independent predictor for mortality, and therapy‐induced regression of LVH improves long‐term outcomes. 8 , 9 , 10 , 12 , 22 Given the lifelong impact that treatment choices have, it is crucial to obtain highly reliable information concerning HMOD, including LVH. Therefore, we strongly recommend implementing standard echocardiography in HMOD screening in hypertensive patients.
This study's assessment of comparing ECG and echocardiography for LVH screening in a hypertension screening setting provides valuable information. Currently, national and international guidelines suggest performing an ECG in routine workup, rather than echocardiography for LVH screening in patients with hypertension. 1 , 2 , 3 This is likely because echocardiography is not widely available in the primary care setting, even though most patients who would derive benefit from the HMOD evaluation, including echocardiography, are under the care of family physicians. Hence, when clinically relevant, insufficient methods should be avoided and an effort should be made to overcome logistical challenges, both in the primary and in the secondary care setting.
The observation about the discrepancy between positive LVH criteria on ECG and echocardiography can be attributed to the nature of the ECG as a tool that provides insights into the electrophysiological state of the heart, rather than directly assessing myocardial muscle mass. Nevertheless, extensive research in large populations shows that being positive for several LVH ECG criteria is associated with an increased risk of CV events. 23 , 24 , 25 , 26 Hence, it would be short‐sighted to dismiss the value of ECG as a tool in assessing CV risk.
In addition to its primary benefit, echocardiography provides supplementary information about cardiac abnormalities besides LVH, thereby yielding additional prognostic value. The findings of the current study indicate a substantial prevalence of valve insufficiency and heart failure among the examined patients. The latter being especially relevant given the emergence of new therapeutic options, such as SGLT2 inhibitors, for managing diastolic heart failure. 27 , 28 , 29 While not identified in this specific population, echocardiography can effectively detect aortic valve stenosis, a potential cause for LVH. Similarly, echocardiography serves well in identifying a bicuspid aortic valve, which is associated with coarctation of the aorta—an infrequent yet well‐established cause of treatment‐resistant hypertension.
In conclusion, the place for ECG in the evaluation of LVH in hypertensive patients is questionable. When solely relying on ECG to diagnose LVH, a substantial part of the patient population will be either inadequately or excessively treated with antihypertensive drugs. Ideally, echocardiography is made easily accessible for those patients where the presence of HMOD adds to treatment decisions. At least for young hypertensive patients, patients with high‐normal blood pressure and therapy‐resistant hypertensive patients in whom HMOD is not otherwise detected.
AUTHOR CONTRIBUTIONS
Marijn Marc Bult: Data analysis and manuscript preparation; including review and editing. Thomas Flint van de Ree: Data curation; analysis and manuscript preparation. Anna Maria Wind: Conception and design of the study; supervision; manuscript editing. Kai Morris Hurley: Data curation and analysis; ensuring the quality of English language usage. Marcel Allard van de Ree: Conception and design of the study; manuscript editing and overall supervision.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflicts of interest.
PATIENT CONSENT STATEMENT
Patient data were anonymized and de‐identified before analyses, as such, the need for written informed patient consent was waived by the institutional review board.
PERMISSION TO REPRODUCE MATERIAL FROM OTHER SOURCES
Not applicable.
CLINICAL TRIAL REGISTRATION
Not applicable.
ACKNOWLEDGMENTS
There was no source of funding
Bult MM, van de Ree TF, Wind AM, Hurley KM, van de Ree MA. The use of echocardiography compared to electrocardiogram when screening for left ventricular hypertrophy in hypertensive patients: A cross‐sectional study. J Clin Hypertens. 2024;26:977–985. 10.1111/jch.14867
DATA AVAILABILITY STATEMENT
Data can be made available upon revision.
REFERENCES
- 1. Williams B, Mancia G, Laurent S, et al. ESC/ESH Guidelines for the management of arterial hypertension. Eur Heart J. 2018;39(33):3021‐3104. [DOI] [PubMed] [Google Scholar]
- 2. National Institute for Health and Care Excellence. Hypertension in adults: Diagnosis and management NG136. Published 2023. Accessed January 7, 2024. https://www.nice.org.uk/guidance/ng136/resources/hypertension‐in‐adults‐diagnosis‐and‐management‐pdf‐66141722710213 [PubMed] [Google Scholar]
- 3. Unger T, Borghi C, Charchar F, et al. 2020 International Society of Hypertension Global Hypertension Practice Guidelines. Hypertension. 2020;75(6):1334‐1357. [DOI] [PubMed] [Google Scholar]
- 4. Gosse P. Left ventricular hypertrophy as a predictor of cardiovascular risk. J Hypertens. 2005;23:S27‐S33. [DOI] [PubMed] [Google Scholar]
- 5. Koren MJ, Devereux RB, Casale PN, Savage DD, Laragh JH. Relation of left ventricular mass and geometry to essential hypertension. Ann Intern Med. 1991;114(5):345‐352. [DOI] [PubMed] [Google Scholar]
- 6. Casale PN, Devereux RB, Milner M, et al. Value of echocardiographic measurement of left ventricular mass in predicting cardiovascular morbid events in hypertensive men. Ann Intern Med. 1986;105(2):173‐178. [DOI] [PubMed] [Google Scholar]
- 7. Krumholz HM, Larson M, Levy D. Prognosis of left ventricular geometric patterns in the Framingham Heart Study. J Am Coll Cardiol. 1995;25(4):879‐884. [DOI] [PubMed] [Google Scholar]
- 8. Lonnebakken MT, Izzo R, Mancusi C, et al. Left ventricular hypertrophy regression during antihypertensive treatment in an outpatient clinic (the Campania salute network). J Am Heart Assoc. 2017;6(3):e004152. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Pierdomenico SD, Lapenna D, Cuccurullo F. Regression of echocardiographic left ventricular hypertrophy after 2 years of therapy reduces cardiovascular risk in patients with essential hypertension. Am J Hypertens. 2008;21(4):464‐470. [DOI] [PubMed] [Google Scholar]
- 10. De Simone G, Devereux RB, Izzo R, et al. Lack of reduction of left ventricular mass in treated hypertension: the strong heart study. J Am Heart Assoc. 2013;2(3):e000144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Devereux RB, Wachtell K, Gerdts E, et al. Prognostic significance of left ventricular mass change during treatment of hypertension. J Am Med Assoc. 2004;292(19):2350‐2356. [DOI] [PubMed] [Google Scholar]
- 12. Muiesan ML, Salvetti M, Monteduro C, et al. Left ventricular concentric geometry during treatment adversely affects cardiovascular prognosis in hypertensive patients. Hypertension. 2004;43(4):731‐738. [DOI] [PubMed] [Google Scholar]
- 13. Bacharova L, Schocken D, Estes EH, Strauss D. The role of ECG in the diagnosis of left ventricular hypertrophy. Curr Cardiol Rev. 2014;10(3):257‐261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Pewsner D, Jüni P, Egger M, Battaglia M, Sundström J, Bachmann LM. Accuracy of electrocardiography in diagnosis of left ventricular hypertrophy in arterial hypertension: systematic review. Br Med J. 2007;335:711. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Levy D, Labib SB, Anderson KM, Christiansen JC, Kannel WB, Castelli WP. Determinants of sensitivity and specificity of electrocardiographic criteria for left ventricular hypertrophy. Circulation. 1990;81(3):815‐820. [DOI] [PubMed] [Google Scholar]
- 16. Keith NM, Wagener HP, Barker NW. Some different types of essential hypertension: their course and prognosis. Am J Med Sci. 1974;268(6):336‐345. [DOI] [PubMed] [Google Scholar]
- 17. Miceli F, Presta V, Citoni B, et al. Conventional and new electrocardiographic criteria for hypertension‐mediated cardiac organ damage: a narrative review. J Clin Hypertens. 2019;21(12):1863‐1871. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Ang DSC, Lang CC. The prognostic value of the ECG in hypertension: where are we now? J Hum Hypertens. 2008;22:460‐467. [DOI] [PubMed] [Google Scholar]
- 19. Marwick TH, Gillebert TC, Aurigemma G, et al. Recommendations on the use of echocardiography in adult hypertension: a report from the European Association of Cardiovascular Imaging (EACVI) and the American Society of Echocardiography (ASE). Eur Heart J Cardiovasc Imaging. 2015;16(6):577‐605. [DOI] [PubMed] [Google Scholar]
- 20. Nagueh SF, Smiseth OA, Appleton CP, et al. Recommendations for the evaluation of left ventricular diastolic function by echocardiography: an update from the American society of echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr. 2016;17(12):1321‐1360. [DOI] [PubMed] [Google Scholar]
- 21. Hinton TC, Adams ZH, Baker RP, et al. Investigation and treatment of high blood pressure in young people: too much medicine or appropriate risk reduction? Hypertension. 2020;75(1):16‐22. [DOI] [PubMed] [Google Scholar]
- 22. Taylor HA, Penman AD, Han H, et al. Left ventricular architecture and survival in African‐Americans free of coronary heart disease (from the atherosclerosis risk in communities [ARIC] study). Am J Cardiol. 2007;99(10):1413‐1420. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Kannel WB, Gordon T, Castellia WP, Margolis JR. Electrocardiographic left ventricular hypertrophy and risk of coronary heart disease: the Framingham study. Ann Intern Med. 1970;72(6):813‐822. [DOI] [PubMed] [Google Scholar]
- 24. Rautaharju PM, Lacroix AZ, Savage DD, et al. Electrocardiographic estimate of left ventricular mass versus radiographic cardiac size and the risk of cardiovascular disease mortality in the epidemiologic follow‐up study of the first national health and nutrition examination survey. Am J Cardiol. 1988;62(1):59‐66. [DOI] [PubMed] [Google Scholar]
- 25. Iribarren C, Round AD, Lu M, Okin PM, McNulty EJ. Cohort study of ECG left ventricular hypertrophy trajectories: ethnic disparities, associations with cardiovascular outcomes, and clinical utility. J Am Heart Assoc. 2017;6(10):e004954. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Lehtonen AO, Puukka P, Varis J, et al. Prevalence and prognosis of ECG abnormalities in normotensive and hypertensive individuals. J Hypertens. 2016;34(5):959‐966. [DOI] [PubMed] [Google Scholar]
- 27. Zhou H, Peng W, Li F, et al. Effect of sodium‐glucose cotransporter 2 inhibitors for heart failure with preserved ejection fraction: a systematic review and meta‐analysis of randomized clinical trials. Front Cardiovasc Med. 2022;9:1‐14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Anker SD, Butler J, Filippatos G, et al. Empagliflozin in heart failure with a preserved ejection fraction. N Engl J Med. 2021;385(16):1451‐1461. [DOI] [PubMed] [Google Scholar]
- 29. Vaduganathan M, Docherty KF, Claggett BL, et al. SGLT‐2 inhibitors in patients with heart failure: a comprehensive meta‐analysis of five randomised controlled trials. Lancet. 2022;400(10354):757‐767. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data can be made available upon revision.
