ABC | Volume 113, Nº5, November 2019

Original Article Saraiva et al. Normal values for new echocardiographic techniques Arq Bras Cardiol. 2019; 113(5):935-945 Ageing correlated with worsening of LA reservoir and conduit function as demonstrated by positive correlation with minimum LA and pre-A volumes and negative correlation with total and passive LA emptying fractions, LAScd, and LASr. Others also found that age correlated positively with minimum LA volume and negatively with total LA emptying fraction 13 or that all LA volumes increased with age. 14 Others also found that LAScd, and LASr decreased with age. 7 The correlation found with age may be partly explained by the known effect of age on LV diastolic function. On multiple regression analysis, age remained the most common covariate associated with 3DE and LA ε parameters. Male gender was also positively associated with LA pre-A volume and negatively associated with total and passive LA emptying fractions. Others found higher maximum and minimum 3DE LA volumes in men, 13 while others found no gender influence on 3D LA volumes. 14 Passive LA emptying fraction, LASr, and LAScd decreased with BMI, demonstrating a possible effect of overweight on LA reservoir and conduit function. In fact, others found that total and passive LA emptying fractions and LA ε decreased with BMI. 19 The 3DE LV volumes described by us are similar to those described previously, 20,21 but larger than those described by others. 22,23 These differences may be attributed to different ethnic backgrounds, as the first two studies are composed of white patients and the last two, patients with different backgrounds. The 3D LV ejection fraction described by us is similar to the one described in those articles. We found larger LV volumes in men, as in others. 20,21 However, after multiple regression analysis, only age was related to 3D LV volumes. We described a negative correlation between age and LV volumes, as previously described. 21,22 The global LV ε found in our population did not differ from previous studies that analyzed LV-GLS, 24-28 LV-GCS 25,26,28 and LV-GRS. 25-28 However, other studies described higher LV-GCS in normal subjects than us 27 and a recent meta-analysis of 24 studies found mean LV-GLS and LV-GRS values very similar to our findings, but higher mean LV-GCS than we did. 29 We found that LV-GCS was higher in women and a tendency for LV-GLS to be higher in women, which is consistent with others that found higher LV-GCS 25,27 and LV-GLS 26,27 in women. Multiple regression analysis confirmed the positive independent relationship between women and absolute LV-GCS values. Although others found correlation between LV ε and age, 25,28 we could not confirm those results. Kocabay et al. 26 did not find any difference in LV ε between different age groups either. 26 Multiple regression analysis demonstrated that LV-GLS absolute values decreased with BMI, as previously described. 26 The LV rotation and twist described by us are similar to the ones previously described, 30 but lower than the ones described by others, 26,31 and higher than the ones described elsewhere. 28 LV twist and torsion increased with age, as described by others. 26,28 Reasons for differences found in mean LV ε and torsion between our study and others may be due to heterogeneity introduced by different machine vendors, 32 age, 25 gender distribution 25-27 and the technique used to measure LV-GLS and LV-GRS. 27 Regarding RV strain, the RV-fwLS described by us was similar to that found in previous studies, 33 but slightly lower than that described by others. 34 Although we did not find correlation between RV ε and age after multiple regression analysis or gender-based differences, others 33 found that RV‑fwLS decreased with ageing, and was lower among men. After multiple regression analysis, RV-GLS absolute values decreased with BMI and ethnicity and RV-fwGLS also decreased with BMI. Although we cannot compare these results to the ones found in the literature, LV-GLS is described to decrease with BMI 26 and ethnicity may influence LV ε . 25 More studies in subjects without known diseases are needed to validate reference values. Limitations Our data was obtained using a machine from a single vendor but the technology of 3DE measurements and STE may differ between different vendors and Echopac software versions. In fact, LV-GLS measured with seven different brands presented a small but significant difference. 35 We used a multi-beat approach to acquire 3D volumes. Multi-beat 3D images usually offer better temporal resolution than single-beat images, but may have lower spatial resolution due to stitching artifacts. 36 However, the correlation with volumes measured by cardiac resonance was described to be excellent, regardless of the number of cardiac cycles used. 37 Recently, newdedicated software tomeasure 3D LA volumes was developed 38 and may result in slightly different data than ours. However, we described good inter- and intra-observer agreements for our 3D LA volume measurements 5 and the correlation between non-dedicated 3D software to measure LA volume and cardiac resonance was described as significant. 39 Our paper did not include enough individuals from all ethnic groups that compose the Brazilian population, such as Asian-Brazilians and native Indians. Another limitation is that we did not perform lab work to exclude patients who could be unaware of associated comorbidities. Also, we did not exclude patients who were overweight. In fact, 41 patients (53%) were overweight (BMI ≥ 25 and < 30 kg/m 2 ). However, overweight is becoming increasingly prevalent among Brazilians and the total percentage in this article is close to the value disclosed by IBGE (56.9%). 40 Conclusions Values for new echocardiographic indexes in normal Brazilian volunteers and their differences between men and women as well as their correlation with age are presented. However, we found differences in reference values between our study and others, which may be due to heterogeneity introduced by different machine vendors, age, gender distribution, ethnicity and the technique used to measure LV, LA or RV ε . Author contributions Conception and design of the research: Saraiva RM, Costa AR. Acquisition of data: Saraiva RM, Scolin EMB, Pacheco NP, Bouret ME, Holanda MT. Analysis and interpretation of the data: Saraiva RM, Bouret ME, Mediano MFF, Holanda MT, Costa AR. Statistical analysis: Saraiva RM, Mediano MFF, Costa AR. Obtaining financing: Saraiva RM. Writing of 942