Malnutrition is a serious health problem that causes over 3.1 million child deaths annually in low- and middle-income nations [1].

All bodily systems, including the cardiovascular system, experience anatomical and functional alterations as a result of malnutrition. Not enough research has been done on how the severity and type of malnutrition affect heart functioning [13]. Whether heart problems are the main conditions linked to malnutrition or not has not been established [3].

This study aimed to report cardiovascular changes in children with moderate and severe acute malnutrition utilizing 12-lead ECG, conventional echo, and 2D speckled echo. The novel part of this study is using speckled echo tracking and 3D volumetric indices for evaluation of subtle cardiac changes that cannot be detected by conventional echo.

Demographic data of the studied cases showed that there are many social and economic factors that played an integral role in the development of malnutrition in those cases. Residing in rural areas was found to be a major determinant of malnutrition. Low levels of paternal education and low-income levels were found to be major risk factors as well. This affects the mother’s awareness of the importance of healthy habits and lifestyle.

One of the most important risk factors of malnutrition, especially among SAM, found was bottle feeding initiated at any age where the majority of the patients had bottle feeding, either alone or combined with breastfeeding. Poor bottle hygiene or improper formula preparation could be the cause, which could result in acute infective diarrheal sickness and other illnesses [14]. Another important risk factor detected was early or late introduction of complementary feeding (CF) as well as poor quality CF. Most SAM patients receive CF after the age of 6 months.

The previous risk factors for malnutrition were agreed upon in previous studies addressing several nutritional risk factors predisposing to child malnutrition [15,16,17,18,19,20].

As we noted, the largest proportion of our cases had acute illness either during the study or 2 weeks apart from performing the study, and nearly half the cases were hospitalized. Gastroenteritis was the most common illness. According to Chelo et al., SAM was more common in people who were not exclusively breastfed and in those who consumed insufficient amounts of calories and protein. They also discovered that hospitalization and the existence of existing medical conditions were contributing risk factors for SAM [21].

All of the present study cases had regular sinus rhythm during the study. They had significantly higher heart rate than controls which can be attributed to many factors including anemia, sepsis/fever, and dehydration. Prior studies have noted regular sinus rhythm, which is consistent with our findings [3, 4].

Low-voltage ECG was predominant in SAM cases with edema. In agreement, previous studies reported that for every 1 g/L elevation in serum albumin concentration, voltage elevated by 0.2 mV [4, 22]. Instead of pericardial effusion, they attributed this voltage drop to the LV’s diminished muscle mass [4]. We could detect pericardial effusion in only one patient, and it was minimal in amount without any hemodynamic significance.

Pathological T-wave changes were evident among the cases (in the form of inverted or flat T waves) and were significantly correlated to hypokalemia and hypophosphatemia. Similar findings have been noted by Brent et al., and Kumar et al., who reported flat or inverted T-wave changes where the T wave returned to normal after treatment with a diet rich in calories and proteins [22, 23]. No study denoted the significance of T-wave changes in children with malnutrition. Contrary to Brent et al., U waves could be detected in our study in different limb leads, V1–V5, with hypokalemia as the predominant electrolyte disturbance [22].

In contrast to previous studies, we could not find a significant difference in QTc between cases and controls [3, 23,24,25]. However, a study in 2019 reported the same findings as ours where they found only 3 patients with prolonged QTc and 24 patients had shortened QTc, out of 88 [22].

However, in agreement with the previous studies, we found significantly greater values of QTc dispersion. The significance of prolonged QTc dispersion is an independent diagnostic marker of apparent life-threatening events (ALTE), such as ventricular arrhythmias and sudden cardiac death [26, 27]. The small sample size in our study or other variables affecting arrhythmia incidence may be the reason we did not detect ventricular arrhythmia. Sudden unexplained death was reported in one of our patients who suffered from a severe form of acute malnutrition (marked wasting); this patient had prolonged QTc and QTc dispersion, and the patient also had septicemia; as we mentioned, we are not yet certain whether death in malnourished patients occurs primarily due to malnutrition or as a part of the original disease as GE or sepsis.

As one could expect, electrolyte disturbances are significantly correlated to ECG changes. Prolonged PR interval was significantly associated with hypophosphatemia, hypomagnesemia, and hyponatremia. Prolonged QTc and QTc dispersion were significantly correlated to hypokalemia. Hypomagnesemia and hypokalemia were markedly correlated with ST segment changes and pathological waves namely U waves.

Regarding the assessment of cardiac systolic function (by conventional echo), there is still no agreement regarding the effect of malnutrition on myocardial function. Brent et al. evaluated LV and RV systolic function using fractional shortening (FS), EF, MAPSE, and TAPSE on 88 Kenyan children with SAM and reported similar results [22]. In another study conducted in Cameroon which utilized TAPSE for evaluation of RV systolic function and found consistent results with ours [21].

The current study evaluated LV and RV systolic function using tissue Doppler imaging by calculating tissue velocity, and no significant changes were found between both groups. TEI index showed no significant difference between both groups, confirming the previous finding of non-affected LV systolic function caused by acute malnutrition. This was in concordance with Brent et al. [22], while others have found significantly affected MPI [20]. TEI index has a limiting factor being affected by heart rate and loading conditions, including preload and afterload.

The current study found a significantly reduced LVEDD and IVS thickness in cases with normal posterior wall thickness. There was a markedly diminished LV mass in cases than controls, but when LV mass was indexed to body surface area (BSA), there was no significant difference between both groups, and these findings are in agreement with previous literature [3, 7, 13, 20, 24, 25, 28].

As per Öcal et al., individuals with PEM had diminished muscle mass due to decreasing myocardial tissue, as seen by the diminished thickness of the left ventricular septum and posterior wall, even though the left ventricular end-systolic and end-diastolic dimensions were within the normal range [7]. These findings lead us to believe that in PEM patients, the heart cannot escape the atrophy that affects other organs, and that the decrease in cardiac mass is proportionate to the decrease in total body mass. Reduced LV mass (LVM) was significantly related to low-voltage ECG.

3D volumetric measurements of LV and RV and speckled echo tracking were utilized as well for estimation of systolic myocardial systolic function. In concordance with M-mode measurements, the current results revealed a marked decrease in both RV and LV systolic and diastolic volumes. However, these findings might not clearly reflect myocardial mass or function, and they may be affected by volume status and ventricular filling at the time of examination as well. Also, 3D LV and RV EF and FAC showed no significant difference.

Strain imaging echo, including LV global longitudinal strain (GLS), is now widely considered a better index of myocardial dysfunction than LV EF and can detect early and subtle myocardial dysfunction when other parameters appear to be normal or inconsistent [29]. Utilizing strain imaging echo for assessment of RV and LV systolic function was an integral part of our study; here, we reported a significant reduction in LV and RV GLS between both groups, even though conventional echo showed no systolic dysfunction in the malnourished group. This denotes that there is subtle myocardial dysfunction that needs sensitive and advanced techniques of measurement, such as strain imaging. Thereby, we recommend using strain echo to detect early myocardial changes in cases of malnutrition. Only one study utilized strain imaging and came up with the same results as ours regarding markedly reduced LV GLS, but RV GLS was not assessed [24].

LV diastolic dysfunction was recorded in this research. More specifically, an elevated average E/e′ ratio at the lateral wall, which is a correlate of myocardial relaxation and LV filling pressure, and a significantly reduced E velocity result in a much lower mitral E/A, suggesting diastolic cardiac issues. This was in keeping with two studies which reported the same finding as ours [21, 30]. Other studies which evaluated only E/A ratio without TDI could not detect any diastolic dysfunction [13, 25, 28]. In a study by Öcal et al., diastolic dysfunction was found in malnourished cases compared to controls, and it might be explained by disorders of energetic metabolism and the predominant tachycardia in malnourished patients [7].

We could not find significant differences between MAM and SAM groups apart from lower voltage ECG and reduced RV function, as noted by TAPSE and 3D RV EF, in SAM cases. However, by reviewing past studies, no significant differences could be detected between both groups, meaning that acute malnutrition as a uniform entity affects the myocardium regardless of moderate or severe form, and our results of RV dysfunction in SAM cases are left for deeper study and on a larger scale of patients.

Prior evidence showed that there is evident myocardial systolic dysfunction in patients with edematous malnutrition (Kwashiorkor) [31, 32], but we could not find significant myocardial dysfunction between edematous and non-edematous patients by measuring EF and FS or even reduced muscle mass. However, our findings were in accordance with two prior studies [7, 20]. We could find that the edematous group had significantly prolonged PR interval, pathological T waves, and, most importantly, significantly reduced LV GLS, denoting that edema (hypoalbuminemia) is an important indicator of subtle or early myocardial dysfunction.

Limitations of the present study include the lack of investigation of the impact of biochemical and clinical complications (most importantly sepsis) in sick children with SAM and the lack of assessment of the measured parameters after nutritional rehabilitation.