Background Moderate acute malnutrition (MAM), defined as a weight-for-height z-score between -2 and -3 and/or a mid-upper arm circumference < 125mm and ≥ 115mm, affects an estimated 33 million children worldwide. In humanitarian emergencies, MAM is treated with food supplements either in the form of corn-soy-blends (CSBs) or lipid based nutrient supplements (LNS). However, there is currently no consensus on the best product and more evidence is needed regarding the composition of these products. Studies investigating supplements for MAM have usually focussed on anthropometric outcomes and little is known about their impact on micronutrient deficiencies such as iron deficiency (ID) and resulting anaemia, which are likely to accompany MAM. Assessing the impact of supplements on ID firstly requires estimating prevalence of ID, which is a challenge because biomarkers of ID, such as serum ferritin (SF), are affected by inflammation. SF needs to be adjusted for inflammation but it is not clear how best to do this. Furthermore, little is known about morbidity in children with MAM. An understanding of the prevalence of infection and inflammation in children with MAM before and after treatment is important not only because it complicates diagnosis of ID but also because it is an important cause of anaemia, may affect response to treatment and because the safety of iron and micronutrient supplements has been questioned particularly in settings where malaria is common. The objectives of this PhD study were to determine prevalence of inflammation, infection, ID and anaemia in children with MAM, to investigate the use of regression models to adjust iron status biomarkers for inflammation and to determine the impact of supplements for MAM treatment on iron status, anaemia and inflammation. Methods A randomised 2x2x3 factorial trial was conducted in Burkina Faso. Children aged 6-23 months with MAM received supplementary foods in the form of LNS or CSB (500kcal/day) referred to as the matrix, containing either dehulled soy (DS) or soy isolate (SI) and different quantities of dry skimmed milk (0, 20 or 50% of total protein). The trial was double blinded with regard to quality of soy and quantity of milk, but not matrix (CSB vs LNS). Haemoglobin (Hb), serum ferritin (SF), serum soluble transferrin receptor (sTfR), serum C-reactive protein (CRP) and serum α1-acid glycoprotein (AGP) were measured at inclusion and after supplementation. Furthermore, morbidity data were collected during clinical examinations carried out by nurses and maternal morbidity recalls. Results Between September 2013 and August 2014 1609 children were enrolled and 61 (3.8%) were lost to follow-up. At baseline, 38% of mothers reported that their children had been ill in the previous two weeks. Furthermore, 71.8% of children had a symptom or infection identified during the physical examination and 24.2 and 66.4% of children had elevated CRP and AGP, respectively. Of asymptomatic children, 10.7 and 46.5% had elevated CRP and AGP, respectively. While symptoms and infections identified were associated with levels of CRP and AGP, they only explained a small amount of their variation as demonstrated by an R2 below 0.2 in all tested ANCOVA models. Prevalence of anaemia and elevated serum soluble transferrin receptor (sTfR) was 70% and 83%, respectively. Due to the high prevalence of inflammation SF needed to be adjusted for inflammation. Linear regression models were identified as a suitable method for the adjustment. After adjustment prevalence of low SF at baseline was 38%. There was an effect of LNS vs CSB on Hb (2 g/L, 95% CI: 1; 4), SF adjusted for inflammation (4.2 μg/L, 2.9; 5.5), sTfR (-0.9 mg/L, -1.3; -0.6) and CRP (0.8 mg/L, 0.4; 1.2). There was no effect of soy quality and milk protein content on Hb, SF adjusted for inflammation, sTfR and CRP. There was no impact of any of the factors on malaria. Conclusion Morbidity, anaemia and iron deficiency among children with MAM in this setting are common. Some children had elevated CRP and/or AGP in the absence of identified symptoms or infections and morbidity data collected only explained a small amount of the variation in CRP and AGP, indicating a presence of subclinical inflammation. The high levels of morbidity in children with MAM indicate that there may be a need for more emphasis of identification and treatment of infections in children with MAM. Supplementation with LNS compared to CSB led to better Hb and iron status, but overall prevalence of anaemia remained high. There appeared to be no benefit of increasing milk content or replacing DS with more expensive SI on iron status and anaemia. Furthermore, the causes and consequences of subclinical inflammation as well as the higher concentrations of acute phase proteins reported in children who received LNS require further investigation.