Abstract

Background: Malnutrition underlies 50% of paediatric morbidity and mortality in sub-Saharan Africa. It is important to look for the underlying causes of the malnutrition, and some clinicians have assumed that the presence of a pericardial effusion indicates underlying tuberculosis (TB). We wished to see how common pericardial effusions are in malnourished children and how their presence or size is related to peripheral oedema or the type of malnutrition of the child, HIV status or to underlying TB.

Methods: We prospectively studied a cohort of children at a regional nutritional rehabilitation unit in Malawi. Echocardiography on admission and follow-up 4 weeks later was performed. During this interval children received therapeutic feeding and any other required medical care. The children were grouped into group 1 (marasmus), group 2 (marasmus with TB), group 3 (marasmic kwashiorkor), group 4 (marasmic kwashiorkor with TB), group 5 (kwashiorkor) and group 6 (kwashiorkor with TB).

Results: Of the 89 children who were enrolled, 28 were marasmic (eight also had TB), 29 had marasmic kwashiorkor (six with TB) and 32 had kwashiorkor (four with TB). In all the children who had a pericardial effusion, its size was greatest at presentation. The overall reduction in pericardial effusion size after 4 weeks of nutritional therapy was significant (2.9 mm change, range 0 to 8.4 mm, p = 0.002). The greatest change in pericardial effusion size was in the children with most peripheral oedema compared with those with no oedema (2.7 mm versus 1.0 mm, p = 0.017).

Conclusions: In severely malnourished children pericardial effusions are common, larger in children with peripheral oedema and respond to nutritional therapy alone.