Despite a significant reduction in the incidence of childhood pneumonia due to the roll-out of immunization, uptake of HIV prevention and treatment and some reduction in the major risk factors (malnutrition, non-exclusive breastfeeding, crowding, and indoor air pollution) pneumonia remains a major public health challenge, particularly in Africa (1–3). Moreover, the reduction of pneumonia mortality lags behind other common pediatric infectious diseases in lower to middle-income countries (LMICs) (2). In 2021, over 37 million pneumonia episodes were reported among children under 5 years, with an incidence rate of 5,750 per 100,000 (1). The majority of pneumonia cases occur in LMICs (1, 3), where undernutrition is also persistently prevalent (4). Undernutrition in the form of wasting, stunting, or being underweight is one of the top ten risk factors for the disease burden among children under 5 years (1). Undernutrition significantly increases the risk of developing severe pneumonia among children (5). Undernutrition also results in a higher frequency of adverse outcomes including mortality (6–9), deterioration in nutritional status (10–13), recurrent infections, and repeated hospital readmissions (14, 15) among children.

Pneumonia is the commonest infectious cause of mortality in children, accounting for over 14% of deaths among children under 5 years, and approximately 740,180 children in 2019 (3, 16). Children with pneumonia may die directly as a result of complications of the disease itself such as hypoxemia (17, 18), or indirectly via secondary infections related to immune compromise (19). Pneumonia also increases the risk of deterioration of children’s nutritional status via increased metabolic demands to support physiological functions such as hyperthermia and increased work of breathing at a time when the children have reduced appetite and are unable to feed adequately (20, 21). Deterioration in nutritional status further compromises their immunity with resultant increased risk of infections, re-admission to hospital, and potentially death (10, 11, 14, 15, 21).

The World Health Organization (WHO) under the Integrated Global Action Plan for the Prevention and Control of Pneumonia and Diarrhea (GAPPD) recommends “continued feeding” in the “diagnose and treat” element (22). However, the nutritional requirements in a child with pneumonia are likely to be higher than those of a well-child (10, 11), and if not addressed could increase the risk of deterioration in nutritional status following the pneumonia episode. Children recovering from pneumonia are likely to require more energy to recover from deficits resulting from increased metabolic demands and anorexia during the illness.

According to the joint consultation of the WHO, United Nations Children’s Fund, World Food Program, and the United Nations High Commissioner for Refugees in October 2008, the desirable nutrient intakes for recovery from MAM are probably in the range between the recommended nutrient intakes for well-nourished children and the intakes recommended in the recovery phase for those with severe malnutrition (23, 24). The options available in Africa to meet these energy requirements are ready-to-use supplementary food (RUSF) and Ready-to-Use Therapeutic Food (RUTF). While RUSF is only available for community nutritional programs, RUTF is used for the follow-on nutritional rehabilitation of children hospitalized with severe acute malnutrition. Previous studies suggest that the provision of one sachet per day (500 kcal) for children under 5 years with moderate malnutrition (defined as a mid-upper arm circumference (MUAC) 11.5 cm − < 12.5 cm) could provide over 49% of the daily energy requirements of children with MAM (25), and improve outcomes (26). Considering the actual nutrient requirement for children with severe pneumonia population is unknown, there is no macronutrient intervention uniquely designed for them. However, nutritional requirements during and after recovery from a pneumonia episode are likely to lie between the daily recommended allowance for age and the requirement for children with moderate malnutrition. RUTF could therefore be extended to children at high risk of undernutrition/malnutrition, such as children recovering from severe pneumonia.

RUTF is a peanut-based paste, specifically designed for children with malnutrition, which provides additional energy, protein, fat and micronutrients. RUTF has been used safely for the treatment of moderate acute malnutrition (MAM) with significantly higher patient recovery, reduced non-recovery, and better weight gain when compared to patients treated with blended foods (27, 28). When compared to nutritional counseling or micronutrient supplementation only, RUTF is associated with better anthropometric outcomes (29). A short course of RUTF has also been considered for African children with acute infections to improve growth with varying results in two cohorts (30, 31). The Children’s Oxygen Administration Strategies-Nutrition Trial (COAST-Nutrition) was designed to investigate whether supplementary feeding with ready-to-use therapeutic food (RUTF) could provide additional nutritional requirements and improve global outcomes in children hospitalized with severe pneumonia across the anthropometric spectrum (32). The trial showed that 90-day mortality and the increment in mid-upper-arm circumference (MUAC) were similar among the children who received RUTF compared to those who did not receive it.

We hypothesized that 1 sachet of RUTF given daily for 8 weeks in addition to the usual diet to undernourished children aged 6–59 months with severe pneumonia would reduce the risk of adverse outcomes (mortality, deterioration in nutritional status, and readmission). If shown to benefit children with undernutrition following admission with pneumonia, RUTF could potentially be rapidly deployed for nutritional support in children with pneumonia.

We included 741 undernourished participants aged 6–59 months; 374 in the intervention (RUTF) arm and 367 in the control (usual diet) arm (Figure 1). Overall, the median age was 16 (IQR 9.26) months and the majority of participants were male (n = 429, 57.9%). The median baseline weight-for-age z score was −1 (IQR – 2.0), height-for-age z score −1 (IQR – 2.0), and weight-for-height z score −0 (IQR – 1.0). The median MUAC was 14 cm (14, 16), and the majority (714, 96.4%) had MUAC ≥12.5 cm. The majority of the participants were afebrile (402, 54.3%), hypoxemic with median oxygen saturation of 89% (87.90), tachypneic (672, 90.7%), and alert (701, 94.6%). The majority of the participants’ caretakers reported difficulty in breathing at baseline (727, 98.1%), and that they had received oral antibiotics or injections before presentation to the hospital (399, 54.9%). On examination, the majority of the participants had chest indrawing (682, 92.2%), and crepitations (571, 77.1%), while 20% (n = 148) were grunting and 29.4% (n = 217) had an audible wheeze. A quarter of the participants (n = 178) had positive malaria rapid diagnostic tests, and 63 (8.6%) had positive malaria blood films. Eleven participants (1.6%) were HIV positive at admission and 74 (10.4%) were found to have sickle cell disease. The median PET severity score for included participants was 2 (IQR 1.3) with the majority of the participants having radiological evidence of pneumonia (430, 58.8%). Baseline characteristics were equally distributed between intervention and control arms (Table 1).

Overall, the primary outcome could not be assessed among 54 (7.2%) participants (Figure 1). The total number of participants who experienced the primary outcome was 370 (49.9%). There was no difference in the occurrence of the primary outcome between the intervention (184/348, 52.9%) and the control arms (186/339, 54.9%) with an adjusted odds ratio of 0.92 (95% CI 0.68, 1.24), p = 0.572 (Table 2). Adverse outcomes experienced by day 90 were deterioration in nutritional status (48%, n = 318), mortality (3%, n = 23) and readmission to hospital (8%, n = 56). There was no difference between the intervention and control groups when the separate components of the primary outcome at day 90 of follow-up were considered. By day 180 of follow-up, 449 (69.5%) participants had experienced the composite of adverse outcomes: 225/334 (67.4%) in the intervention and 224/312 (71.8%) in the control arm. There was no significant difference in the occurrence of the composite outcome between the intervention and control arms [adjusted odds ratio 0.80 (95% CI 0.57, 1.13)], or in the separate components by day 180 of follow-up (Table 2).

The median age of the participants who experienced adverse outcomes was 13 months (IQR 8, 21), and the majority were male, 253 (56.3%). The median weight-for-age baseline z score for these children was −0 (IQR − 1, 0), height-for-age baseline z score −1 (IQR − 1, 0), weight-for-length/height z-score −0 (IQR − 1, 1) and MUAC 15 cm (IQR 14, 16). There were relatively more children with grunting at baseline who experienced adverse outcomes 101/449 (22.5%) versus 29/197 (14.7%). However, there was generally an equal distribution of characteristics between participants who experienced adverse outcomes and those who did not (Table 3).

The risk of experiencing adverse outcomes reduced with increasing age (adjusted odds ratio 0.53, 95% CI 0.45, 0.62, p < 0.001). Sex and PET score were not significantly associated with experiencing an adverse outcome, (adjusted odds ratio 1.31, (95% CI 0.91, 1.88), p = 0.147 and 1.11, (95% CI 0.92, 1.33), p = 0.282 respectively) (Table 3).

In this secondary analysis of the COAST-Nutrition trial, we specifically investigated the ability of 1 sachet of RUTF given daily for 8 weeks in addition to the usual diet to reduce adverse outcomes including mortality, readmission, and deterioration in nutritional status by day 90 compared to usual diet alone among undernourished children aged 6–59 months admitted for severe pneumonia. We found that a majority (54%) of undernourished children experienced adverse outcomes by day 90 of follow up and up to 70% experienced adverse outcomes by day 180. Supplementing the usual diet with RUTF did not reduce adverse outcomes in this population. The risk of experiencing adverse outcomes reduced with increasing age.

While the impact of RUTF on growth in children recovering from non-severe acute infections has been investigated with variable results (30, 31), there is a paucity of data among children with severe pneumonia. Macronutrient supplementation with modified foods as well as supplementary foods like RUTF have been explored among children with Moderate Acute Malnutrition (MAM), defined as weight-for-height/length > − 3 but ≤ − 2 in LMICs, leading to improved recovery rate from MAM and improved nutritional status (27, 29). Lipid-based nutritional supplements including RUTF did not significantly reduce mortality or progression to severe acute malnutrition among children with MAM (27). However, when compared to counseling alone, approximately 550 kcal/day of supplementary food (such as RUTF) for ≤ 12 weeks among children with MAM in Sierra Leone was protective from severe acute malnutrition and death (28).

Van der Kam et al. (30) showed that RUTF did not reduce the incidence of severe malnutrition during recovery among acutely ill Nigerian children without severe malnutrition, including those with pneumonia. This was thought to be possibly due to the short duration of intervention of 14 days (30). Nevertheless, our study gave the intervention (RUTF) for a longer period (8 weeks), close to that successfully used for MAM in Sierra Leone (28), but report similar results to the Nigerian study. Although the COAST Nutrition patients in this analysis were severely ill and undernourished (27, 3.6% with MAM) RUTF did not reduce adverse outcomes. This suggests that children with more severe forms of undernutrition (defined as a weight-for-age z score, height-for-age z score, or weight-for-height/length z score below −2 compared to the WHO reference population or mid-upper arm circumference (MUAC) below 12.5 cm) might benefit more from supplementation with RUTF as demonstrated by Rajabi et al. (28). It is possible that the nutritional requirements of children with severe pneumonia are not met by RUTF, since this food was specifically designed for severe malnutrition. Alternatively, their risk of adverse outcomes cannot be reversed solely by nutritional supplementation.

About 48% of participants in our study had a deterioration in nutritional status by day 90 of follow-up. This highlights the role of pneumonia as a causative or aggravating agent for undernutrition in children under 5 years. The fact that nutritional supplementation with RUTF did not reduce the risk of adverse outcomes among children with severe pneumonia underscores the need to strengthen primary prevention of undernutrition through exclusive breastfeeding, complementary feeding, and healthy young child feeding to prevent severe infections and adverse outcomes in children. Our finding that adverse events reduced with increasing age is in keeping with findings elsewhere (41, 42), and is probably due to the relatively less experienced immune system of younger children which increases their risk of severe disease.

To our knowledge, this is the first study to assess the effect of RUTF on undernourished children with severe pneumonia (complicated by hypoxemia) in LMICs. However, we acknowledge several limitations. The sample size was estimated to detect a large difference (58%) between the intervention and control arm. The study was therefore not powered to detect clinically significant differences smaller than 58%. Given the nature of the intervention, participants and health workers in the study were not blinded and this may have led to information bias. The participants included in this study were a subgroup of the larger randomized controlled trial population based on specific criteria, which may have introduced selection bias. However, there was no significant difference in baseline characteristics between intervention and control groups. Children with severe malnutrition were excluded from the main trial and this sub-analysis as it would have been unethical to randomize them to nutritional supplementation rather than therapeutic doses of RUTF. Adherence to RUTF, as well as food given as part of the usual diet, were not systematically assessed, so co-intervention may have occurred in the study thus diminishing differences between intervention and control groups. Finally, we were ethically required to give parents nutritional counseling in both arms (which is not routinely done in standard practice). This might have improved the nutritional practices in both arms and dampened the effect of the intervention.

Our study showed that supplementation of the usual diet with ready-to-use therapeutic food (RUTF) for 8 weeks among undernourished children aged 6–59 months admitted for severe pneumonia did not reduce adverse outcomes (mortality, re-admission, and deterioration in nutritional status). Further studies using targeted nutritional supplementation specifically meeting the metabolic demands of under-nourished children aged 6–59 months recovering from severe pneumonia are recommended.