Abstract
Background
Current diagnostic ECG criteria of left ventricular hypertrophy in obese patients are still lacking.
Objective
To assess the current ECG diagnostic criteria of LVH, and to validate our previously proposed criteria in a group of patients with morbid obesity.
Methods
A group of consecutive 429 obese patients (MOP) with BMI of at least 35 kg/m2 (mean age 38.6 ± 8.9 years, BMI 48.7 ± 9.0 kg/m2; 323 females, 106 males) were included.
Results
The diagnosis of LVH in MOPs was confirmed only by RaVL of 7.5 mm, Cornell index of 12.5 mm; Cornell index × QRS duration of 1,125 mm × ms and Romhilt–Estes score of 1. None of the criteria proposed to date is appropriate in super‐morbidly obese patients.
Conclusion
Our study confirmed that none of the currently used voltage‐based ECG criteria is appropriate for diagnosing LVH in morbidly obese patients. Further studies are required.
Keywords: Noninvasive techniques ‐ ECHO < Clinical, Noninvasive techniques ‐ electrocardiography < Clinical
1. INTRODUCTION
Obesity is an established risk factor of wide spectrum of cardiovascular diseases. Despite increasing incidence of obesity, its numerous cardiac and hemodynamic complications are relatively poorly studied (Domienik & Pruszczyk, 2007; Wyatt, Winters, & Dubbert, 2006). Left ventricular hypertrophy (LVH) is a common cardiac complication in patients with morbid obesity (Domienik & Pruszczyk, 2007) and was proven to be an important cardiovascular risk factor (Cuspidi, Rescaldani, Sala, & Grassi, 2014; Domienik‐Karłowicz et al., 2011). Therefore, appropriate diagnosis of left ventricular hypertrophy is a very difficult yet important task in everyday clinical practice (Cuspidi, Sala, & Grassi, 2013; Domienik‐Karłowicz et al., 2011; Soteriades et al., 2011). Importantly echocardiography reveals LVH in about 56% cases (Cuspidi et al., 2014). However, standard ECG is widely available and can detect LVH on general population. Current diagnostic ECG criteria of LVH in obese patients are still lacking. The available literature contains no unambiguous, diagnostic criteria of left ventricular hypertrophy in morbidly obese patients.
1.1. Objective
The objective of our study was to assess the current ECG diagnostic criteria of left ventricular hypertrophy, and to validate our previously proposed criteria in a group of patients with morbid obesity. Moreover, we intended to optimize those diagnostic criteria.
2. METHODS
A group of consecutive 429 morbidly obese patients with BMI of at least 35 kg/m2 (mean age 38.6 ± 8.9 year, BMI 48.7 ± 9.0 kg/m2; 323 females, 106 males) were included in the study group. We did not include patients with a history of myocardial infarction, atrial fibrillation, left or right bundle branch block, significant valvular lesions, or chronic obstructive pulmonary disease. Basic characteristics of the patients are presented in Table 1. All patients were subjected to standard echocardiographic examination as well as transthoracic echocardiographic examination at the same day.
Table 1.
Basic clinical and echocardiographic characteristics of the patients (mean ± SD)
| Women (n = 323) | Men (n = 106) | |
|---|---|---|
| Body mass (kg) | 129.3 ± 22.3 | 155.1 ± 22.07 |
| BMI (kg/m2) | 47.1 ± 6.44 | 49.1 ± 6.5 |
| Hypertension n (%) | 306 (94.7%) | 106 (100%) |
| Diabetes n (%) | 26 (8.04%) | 22 (20.75%) |
| Prediabetes n (%) | 88 (27.24%) | 39 (36.79%) |
| Metabolic syndrome | 322 (99.69%) | 106 (100%) |
| LVEDd (mm) | 47.1 ± 5.1 | 50.2 ± 5 |
| IVS (mm) | 11.3 ± 1.2 | 12.1 ± 1.2 |
| LVPW (mm) | 11.4 ± 1.1 | 11.9 ± 1.2 |
| LVM | 229.6 ± 62.9 | 284.6 ± 65.1 |
2.1. Standard electrocardiography
All electrocardiographic examinations were carried out with a Philips Page Writer Trim II device (Philips, Andover, MA, USA). Electrocardiographic LVH criteria were used according to European Society of Cardiology criteria: RaVL > 11 mm; RV5 > 26 mm; RV6 > 26 mm; Rmax + Smax > 45 mm; SV1 + RV5–6 > 35 mm, Sokolow–Lyon index; RaVL + SV3 > 28 mm in men and >20 mm in women, Cornell criteria; Cornell product >2,436 mm × ms; RI + SIII > 25 mm; SV1 or SV2 ≥ 30 mm and Romhillt–Estes score >5 points (Mancia et al., 2014).
2.2. Standard echocardiography
Transthoracic echocardiographic examinations were performed using a Philips IE 33 system (Philips, Andover, MA, USA) and assessed according to recommendations of European Society of Cardiology. End diastolic left ventricular dimension, intraventricular septum thickness, and posterior wall thickness were measured in late diastole defined by R wave of simultaneous ECG tracing. The left ventricular mass was calculated from Devereux formula: LVM (g) = 1.04[(IVS + LVDD + LVPW)3 − LVDD3] − 13.6. Next, left ventricular mass was adjusted for body mass according to the formula LVMI [g/m2] = LVM/BSA, with LVMI > 110 g/m2 and LVMI > 132 g/m2 being considered the reference values for female and male patients, respectively.
2.3. Statistical analysis
Statistical analysis included descriptive statistics (quantitative data, means ± standard deviations; qualitative data, counts, and percentages). Receiver operating curve analysis was applied to determine the area under the curve (AUC). Optimal cutoff values were selected using standard Youden criterion. We checked sensitivity and specificity of currently accepted ECG criteria. R software (R Core Team, 2014) was used for statistical calculations.
3. RESULTS
The analysis revealed that none of the current criteria for the diagnosis of left ventricular hypertrophy is adequate for diagnosing LVH in morbidly obese patients. The results of the analysis are presented in Table 2.
Table 2.
Assessment of the current criteria for the diagnosis of left ventricular hypertrophy in morbidly obese patients
| Criteria | Sens | Spec | PPV | NPV |
|---|---|---|---|---|
| RaVL > 11 mm | 5.5% | 94.2% | 45.8% | 53.0% |
| RV5 > 26 mm | 0.0% | 99.6% | 0.0% | 52.9% |
| RV6 > 26 mm | 0.0% | 100.0% | 52.7% | |
| RV5 or RV6 > 26 mm | 0.0% | 99.6% | 0.0% | 52.9% |
| RaVL + SV3 > 28 mm in men and >20 mm in women | 4.5% | 99.1% | 81.8% | 54.0% |
| Cornell product >2,436 mm × ms | 0.0% | 99.1% | 0.0% | 52.8% |
| Rmax + Smax > 45 mm | 3.0% | 98.7% | 66.7% | 53.1% |
| SV1 or SV2 ≥ 30 mm | 0.0% | 100.0% | 52.7% | |
| SV1 + RV5–6 > 35 mm | 0.0% | 100.0% | 52.7% |
Results presented in Table 2 are consistent with those previously presented by our team. We analyzed the ROC curves to reassess the diagnostic value of ECG parameters in morbidly obese patients (Domienik‐Karłowicz et al., 2011). In line with the earlier conclusions Cornell indices are clinically significant in this group of patients. Romhilt–Estes score was characterized by AUC value of 0.569, 95% CI 0.523–0.615. RaVL values were also of note. Details are presented in Table 3 which also contains novel electrocardiographic diagnostic criteria in LVH based on the analysis of ROC curves.
Table 3.
Reassessment of the diagnostic value of ECG parameters in morbidly obese patients. Novel electrocardiographic diagnostic criteria in LVH
| Criteria | Sens | Spec | PPV | NPV | AUC |
|---|---|---|---|---|---|
| RaVL > 11 mm | 34.5% | 79.6% | 60.0% | 57.7% | 57.5% |
| RV5 > 26 mm | 3.0% | 98.7% | 66.7% | 53.1% | 45.3% |
| RV6 > 26 mm | 3.5% | 99.6% | 87.5% | 53.5% | 47.2% |
| RV5 or RV6 > 26 mm | 3.5% | 98.7% | 70.0% | 53.6% | 47.0% |
| RaVL + SV3 > 28 mm in men and >20 mm in women | 63.0% | 58.4% | 57.3% | 64.1% | 62.2% |
| Cornell product >2,436 mm × ms | 60.5% | 61.1% | 57.9% | 63.6% | 62.5% |
| Rmax + Smax > 45 mm | 55.2% | 53.5% | 51.6% | 57.1% | 52.8% |
| SV1 or SV2 ≥ 30 mm | 22.2% | 83.2% | 54.2% | 54.3% | 51.6% |
| SV1 + RV5–6 > 35 mm | 3.4% | 98.7% | 70.0% | 53.2% | 46.3% |
| Romhilt–Estes score | 66.5% | 46.5% | 52.7% | 60.7% | 56.9% |
Due to the potential correlations with fat distribution and gender, separate subanalyses were performed in female and male patients. Respective results are presented in Tables 4 and 5.
Table 4.
Assessment of the current criteria for the diagnosis of left ventricular hypertrophy in morbidly obese women
| Criteria (women) | Cutoff | Sens | Spec | PPV | NPV | AUC |
|---|---|---|---|---|---|---|
| RaVL | 7.5 | 35.8% | 84.5% | 69.5% | 57.1% | 61.2% |
| RaVL | 7.5 | 35.6% | 84.6% | 69.5% | 57.1% | 61.1% |
| RV5 | 16.5 | 5.0% | 96.9% | 61.5% | 50.5% | 43.6% |
| RV6 | 17 | 4.4% | 100.0% | 100.0% | 51.1% | 45.7% |
| RV5 or RV6 | 16.5 | 5.7% | 96.9% | 64.3% | 51.1% | 45.9% |
| Cornell index | 12.5 | 61.6% | 68.5% | 65.8% | 64.5% | 66.9% |
| Cornell Criteria × QRS duration | 1,120 | 60.4% | 70.4% | 66.7% | 64.4% | 67.4% |
| Rmax + Smax | 11.5 | 59.0% | 54.3% | 56.2% | 57.1% | 54.4% |
| SV1 or SV2 | 9.5 | 41.6% | 66.0% | 54.9% | 53.2% | 53.7% |
| SV1 + RV5 or SV6 | 16.5 | 5.6% | 96.9% | 64.3% | 50.8% | 45.3% |
| Romhilt–Estes score | 1 | 65.2% | 53.7% | 58.3% | 60.8% | 59.9% |
Table 5.
Assessment of the current criteria for the diagnosis of left ventricular hypertrophy in morbidly obese men
| Criteria (men) | Cutoff | Sens | Spec | PPV | NPV | AUC |
|---|---|---|---|---|---|---|
| RaVL | 6 | 56.1% | 51.6% | 42.6% | 64.7% | 48.6% |
| RV5 | 9.5 | 56.1% | 50.0% | 41.8% | 64.0% | 48.7% |
| RV6 | 10 | 34.1% | 71.9% | 43.8% | 63.0% | 47.8% |
| RV5 or RV6 | 9.5 | 56.1% | 50.0% | 41.8% | 64.0% | 48.6% |
| Cornell index | 11 | 87.8% | 20.3% | 41.4% | 72.2% | 53.4% |
| Cornell Criteria × QRS duration | 1,530 | 41.5% | 75.0% | 51.5% | 66.7% | 53.7% |
| Rmax + Smax | 14 | 40.5% | 75.0% | 51.5% | 65.8% | 49.6% |
| SV1 or SV2 | 14 | 23.8% | 89.1% | 58.8% | 64.0% | 48.9% |
| SV1 + RV5 or SV6 | 9.5 | 54.8% | 50.0% | 41.8% | 62.7% | 47.5% |
| Romhilt–Estes score | 0 | 100.0% | 0.0% | 39.6% | 49.8% |
Considering the dependence of voltage‐based criteria on the body weight, separate subanalyses were conducted in patients with BMI <50 and >50 kg/m2; however, this subanalysis was not of value.
ROC analysis for Cornell criterion × QRS duration, as well as Cornell index in LVH assessment in morbidly obese patients was particularly assessed (AUC 0.674; 95% CI 0.615–0.732). Respective results are presented in Figures 1 and 2.
Figure 1.
ROC analysis for Cornell criterion × QRS duration in LVH assessment in morbidly obese patients
Figure 2.
ROC analysis for Cornell index in LVH assessment in morbidly obese patients
4. DISCUSSION
A common change observed in morbidly obese patients is the reduced voltage of QRS complexes in both precordial and limb leads (Fraley, Birchem, Senkottaiyan, & Alpert, 2005; Mutiso, Rono, & Bukachi, 2014). This is probably due to the fact that accumulation of subcutaneous adipose tissue, epicardial adipose tissue (EAT), as well as adipose tissue located within the thorax have impact on the reduction of the electric signals received at the body surface (Bakkum et al., 2015). In a straightforward fashion, this implies reduced sensitivity of ECG in the diagnostics of left ventricular hypertrophy (Cuspidi et al., 2014; Domienik‐Karłowicz et al., 2011; Rodrigues et al., 2015).
Echocardiography is the standard method used for diagnosing LVH in obese patients. To date, Cornell index or Cornell product were considered to be the best criteria for the assessment of left ventricular hypertrophy in obese patients (Cuspidi et al., 2014; Domienik‐Karłowicz et al., 2011). Our own research also suggested a limited applicability of Cornell index and Romhilt–Estes score in the diagnosis of LVH. Attempts at validating both the standard and the previously proposed criteria failed. As of this day, according to Cuspidi, we consider that diagnosing LVH on the basis of ECG examinations is a “mission impossible” (Cuspidi et al., 2014, 2013; Domienik‐Karłowicz et al., 2011; Rodrigues et al., 2015).
5. CONCLUSIONS
Our study confirmed that none of the currently used voltage‐based ECG criteria is appropriate for diagnosing LVH in morbidly obese patients. As we proposed before, only RaVL Romhilt–Estes scale and Cornell indices could be helpful for the identification of LVH in the group of MOPs. The diagnosis of LVH in MOPs was confirmed by RaVL of 7.5 mm, Cornell index of 12.5 mm; Cornell index × QRS duration of 1,125 mm × ms and Romhilt–Estes score of 1. None of the criteria proposed to date is appropriate in super‐morbidly obese patients. Due to the spreading epidemic of obesity further studies are required to facilitate detection of LVH by electrocardiographic methods.
CONFLICT OF INTEREST
All authors confirm no conflict of interests.
Domienik‐Karłowicz J, Rymarczyk Z, Lisik W, et al. Questionable validity of left ventricular hypertrophy cutoff values in morbidly and super‐morbidly obese patients. Ann Noninvasive Electrocardiol. 2018;23:e12564 10.1111/anec.12564
REFERENCES
- Bakkum, M. J. , Danad, I. , Romijn, M. A. , Stuijfzand, W. J. , Leonora, R. M. , Tulevski, I. I. , … Knaapen, P. (2015). The impact of obesity on the relationship between epicardial adipose tissue, left ventricular mass and coronary microvascular function. European Journal of Nuclear Medicine and Molecular Imaging, 42(10), 1562–1573. 10.1007/s00259-015-3087-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cuspidi, C. , Rescaldani, M. , Sala, C. , & Grassi, G. (2014). Left‐ventricular hypertrophy and obesity: A systematic review and meta‐analysis of echocardiographic studies. Journal of Hypertension, 32(1), 16–25. 10.1097/HJH.0b013e328364fb58 [DOI] [PubMed] [Google Scholar]
- Cuspidi, C. , Sala, C. , & Grassi, G. (2013). Detection of left ventricular hypertrophy in obesity: Mission impossible? Journal of Hypertension, 31(2), 256–258. 10.1097/HJH.0b013e32835ca135 [DOI] [PubMed] [Google Scholar]
- Domienik, J. , & Pruszczyk, P. (2007). Obesity as a risk factor of cardiovascular complications. Polski Przeglad Kardiologiczny, 9(6), 432–436. [Google Scholar]
- Domienik‐Karłowicz, J. , Lichodziejewska, B. , Lisik, W. , Ciurzyński, M. , Bienias, P. , Chmura, A. , & Pruszczyk, P. (2011). Electrocardiographic criteria of left ventricular hypertrophy in patients with morbid obesity. Annals of Noninvasive Electrocardiology, 16(3), 258–262. 10.1111/j.1542-474X.2011.00440.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fraley, M. A. , Birchem, J. A. , Senkottaiyan, N. , & Alpert, M. A. (2005). Obesity and the electrocardiogram. Obesity Reviews, 6(4), 275–281. 10.1111/j.1467-789X.2005.00199.x [DOI] [PubMed] [Google Scholar]
- Mancia, G. , Fagard, R. , Narkiewicz, K. , Redon, J. , Zanchetti, A. , Böhm, M. , … Task Force for the Management of Arterial Hypertension of the European Society of Hypertension and the European Society of Cardiology (2014). 2013 ESH/ESC practice guidelines for the management of arterial hypertension. Blood Press, 23(1), 3–16. 10.3109/08037051.2014.868629 [DOI] [PubMed] [Google Scholar]
- Mutiso, S. K. , Rono, D. K. , & Bukachi, F. (2014). Relationship between anthropometric measures and early electrocardiographic changes in obese rats. BMC Research Notes, 7, 931 10.1186/1756-0500-7-931 [DOI] [PMC free article] [PubMed] [Google Scholar]
- R Core Team (2014). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; Retrieved from https://www.R-project.org/ [Google Scholar]
- Rodrigues, J. C. , McIntyre, B. , Dastidar, A. G. , Lyen, S. M. , Ratcliffe, L. E. , Burchell, A. E. , … Manghat, N. E. (2015). The effect of obesity on electrocardiographic detection of hypertensive left ventricular hypertrophy: Recalibration against cardiac magnetic resonance. Journal of Human Hypertension, 30, 197–203. 10.1038/jhh.2015.58 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Soteriades, E. S. , Targino, M. C. , Talias, M. A. , Hauser, R. , Kawachi, I. , Christiani, D. C. , & Kales, S. N. (2011). Obesity and risk of LVH and ECG abnormalities in US fire fighters. Journal of Occupational and Environmental Medicine, 53, 867–871. 10.1097/JOM.0b013e318221c6fe [DOI] [PubMed] [Google Scholar]
- Wyatt, S. B. , Winters, K. P. , & Dubbert, P. M. (2006). Overweight and obesity: Prevalence, consequences, and causes of a growing public health problem. The American Journal of the Medical Sciences, 331(4), 166–174. 10.1097/00000441-200604000-00002 [DOI] [PubMed] [Google Scholar]