research proposal on malnutrition in zambia

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 17 April 2024

Malnutrition enteropathy in Zambian and Zimbabwean children with severe acute malnutrition: A multi-arm randomized phase II trial

• Kanta Chandwe 1   na1 ,
• Mutsa Bwakura-Dangarembizi 2 , 3   na1 ,
• Beatrice Amadi 1 ,
• Gertrude Tawodzera 2 ,
• Deophine Ngosa 1 ,
• Anesu Dzikiti 2 ,
• Nivea Chulu 1 ,
• Robert Makuyana 2 ,
• Kanekwa Zyambo   ORCID: orcid.org/0000-0001-5686-5319 1 ,
• Kuda Mutasa 2 ,
• Chola Mulenga 1 ,
• Ellen Besa   ORCID: orcid.org/0000-0001-5606-9435 1 ,
• Jonathan P. Sturgeon   ORCID: orcid.org/0000-0003-1318-1739 2 , 4 ,
• Shepherd Mudzingwa 2 ,
• Bwalya Simunyola 1 ,
• Lydia Kazhila 1 ,
• Masuzyo Zyambo 5 ,
• Hazel Sonkwe 5 ,
• Batsirai Mutasa 2 ,
• Miyoba Chipunza 1 ,
• Virginia Sauramba 2 ,
• Lisa Langhaug   ORCID: orcid.org/0000-0002-9131-8158 2 ,
• Victor Mudenda 1 ,
• Simon H. Murch   ORCID: orcid.org/0000-0002-3870-8229 6 ,
• Susan Hill 7 ,
• Raymond J. Playford   ORCID: orcid.org/0000-0003-1235-8504 8 , 9 ,
• Kelley VanBuskirk   ORCID: orcid.org/0009-0004-9110-9059 1 ,
• Andrew J. Prendergast   ORCID: orcid.org/0000-0001-7904-7992 2 , 4 &
• Paul Kelly   ORCID: orcid.org/0000-0003-0844-6448 1 , 4  

Nature Communications volume  15 , Article number:  2910 ( 2024 ) Cite this article

1120 Accesses

38 Altmetric

Metrics details

• Intestinal diseases
• Paediatric research

Malnutrition underlies almost half of all child deaths globally. Severe Acute Malnutrition (SAM) carries unacceptable mortality, particularly if accompanied by infection or medical complications, including enteropathy. We evaluated four interventions for malnutrition enteropathy in a multi-centre phase II multi-arm trial in Zambia and Zimbabwe and completed in 2021. The purpose of this trial was to identify therapies which could be taken forward into phase III trials. Children of either sex were eligible for inclusion if aged 6–59 months and hospitalised with SAM (using WHO definitions: WLZ <−3, and/or MUAC <11.5 cm, and/or bilateral pedal oedema), with written, informed consent from the primary caregiver. We randomised 125 children hospitalised with complicated SAM to 14 days treatment with (i) bovine colostrum ( n = 25), (ii) N-acetyl glucosamine ( n = 24), (iii) subcutaneous teduglutide ( n = 26), (iv) budesonide ( n = 25) or (v) standard care only ( n = 25). The primary endpoint was a composite of faecal biomarkers (myeloperoxidase, neopterin, α 1 -antitrypsin). Laboratory assessments, but not treatments, were blinded. Per-protocol analysis used ANCOVA, adjusted for baseline biomarker value, sex, oedema, HIV status, diarrhoea, weight-for-length Z-score, and study site, with pre-specified significance of P < 0.10. Of 143 children screened, 125 were randomised. Teduglutide reduced the primary endpoint of biomarkers of mucosal damage (effect size −0.89 (90% CI: −1.69,−0.10) P = 0.07), while colostrum (−0.58 (−1.4, 0.23) P = 0.24), N-acetyl glucosamine (−0.20 (−1.01, 0.60) P = 0.67), and budesonide (−0.50 (−1.33, 0.33) P = 0.32) had no significant effect. All interventions proved safe. This work suggests that treatment of enteropathy may be beneficial in children with complicated malnutrition. The trial was registered at ClinicalTrials.gov with the identifier NCT03716115.

Similar content being viewed by others

Sepsis-associated acute kidney injury: consensus report of the 28th Acute Disease Quality Initiative workgroup

Short-chain fatty acids: linking diet, the microbiome and immunity

The International Scientific Association of Probiotics and Prebiotics (ISAPP) consensus statement on the definition and scope of postbiotics

Introduction.

Despite 17 million annual worldwide cases of childhood severe acute malnutrition (SAM), and its high associated mortality when children are hospitalized with complications 1 , 2 , there have been few new treatments for over three decades for children with complicated SAM. Current management follows steps in the WHO guidelines launched in 1999 3 , 4 , 5 , 6 but there is an acknowledged lack of evidence for many interventions 1 , 7 and consensus that more trials are needed. In sub-Saharan Africa, HIV has had a major impact on SAM, causing higher mortality during 8 , 9 and after admission 2 , 10 , 11 , 12 , 13 , complications like persistent diarrhoea 9 , and prolonged hospital admission.

Even after recovery from acute SAM, there are very common chronic consequences, including both stunting of linear growth and short- and long-term inhibition of neurodevelopmental potential 14 . The effects of chronic malnutrition upon cognitive functioning are particularly notable in the domains of attention, memory and learning, contributing to poor school performance 15 and corresponding with abnormalities on neuroimaging 16 . Such long-term consequences may not be averted by providing improved nutrition alone, as small intestinal absorption of nutrients and essential micronutrients is compromised by ongoing malnutrition enteropathy 17 . Effective treatment of malnutrition enteropathy is thus likely to have major benefits for the development of large numbers of children within resource-poor countries.

The small intestinal mucosal damage now recognised as malnutrition enteropathy was first recognised in SAM in the 1960s 18 , 19 . More recent studies confirm the very high frequency of malnutrition enteropathy in resource poor countries and an association between such gut inflammation and mortality in complicated SAM 20 . There is therefore considerable interest in malnutrition enteropathy, which varies from mild villus blunting and inflammation to a severe state with total villus atrophy similar to coeliac disease. Malnutrition enteropathy is characterised by severe epithelial lesions 21 , 22 accompanied by mucosal inflammation in the epithelium and lamina propria together with depletion of secretory cells 23 . The epithelial lesions permit microbial translocation from the gut lumen driving systemic inflammation 24 , 25 . Transcriptomic analysis of mucosal biopsies confirms links between inflammation, villus blunting, microbial translocation and epithelial leakiness 26 .

Recognising that fresh approaches are needed to ameliorate the mucosal damage that characterises malnutrition enteropathy we evaluated four potential therapies in a multi-arm, phase II, randomised controlled trial in two tertiary hospitals in southern Africa (Lusaka, Zambia and Harare, Zimbabwe) 27 . The Therapeutic Approaches to Malnutrition Enteropathy (TAME) trial tested the hypothesis that one or more of these therapies could reduce the severity of malnutrition enteropathy in children with SAM. Each intervention was chosen because of its potential for enhancing intestinal repair. Bovine colostrum contains abundant growth factors, including insulin-like and epidermal growth factors 28 , and demonstrates efficacy in ulcerative colitis 29 , 30 . N-acetyl glucosamine may restore the intestinal barrier since glycosylation is deficient in SAM 31 , 32 , and it has been shown to promote mucosal healing in inflammatory bowel disease (IBD) 33 . Teduglutide improves nutrient absorption through mucosal regeneration in intestinal failure 34 , 35 . Budesonide suppresses inflammation with minimal systemic exposure in IBD 36 . A broad range of endpoints was chosen to assess several domains of pathophysiology 37 , 38 relevant to restoring mucosal integrity. The primary endpoint, a composite of faecal biomarkers (myeloperoxidase, α1-antitrypsin, and neopterin) was chosen to reflect mucosal inflammation and loss of barrier function. Secondary endpoints were chosen to reflect enterocyte damage (plasma intestinal fatty acid-binding protein) and the systemic response to microbial translocation (lipopolysaccharide-binding protein (LBP), C-reactive protein (CRP), soluble CD163, soluble CD14), alongside anthropometric measures of nutritional recovery, death, adverse events, diarrhoea, fever and recovery from oedema.

The TAME trial was conducted in the Children’s Hospital of University Teaching Hospital, Lusaka, and Sally Mugabe Hospital, Harare. The third planned site (Parirenyatwa Hospital, Harare) was closed to recruitment due to its use as a COVID-19 centre. Children hospitalised with complicated SAM were enrolled once they were considered stable and ready to transition to higher calorie intakes, to avoid anticipated high early mortality in this high-risk population. Of 143 children screened, 133 were eligible: 8 declined, leaving 125 children enrolled: 62 in Lusaka and 63 in Harare (Fig.  1 , Table  1 ), of whom 43% were female (Table  1 ). Causes of ineligibility were weight <5 kg ( n = 1), clinical instability ( n = 1), haemoglobin <6 g/dL ( n = 2), death prior to enrolment ( n = 1), or resolution of SAM ( n = 5). Children were randomised a median of 5.5 (range 1–21) days after admission (Table  2 ), following blood and stool collection to measure baseline biomarkers; no biomarker data were available from earlier in the hospital admission. One child died and two withdrew before day 15, so 122 children (98%) contributed outcome data (Fig.  1 ). Two further children died and one withdrew between the end of treatment and the day 28 follow-up visit. Adherence and completion were very high: 122/124 (98%) children who survived to day 15 received all planned doses.

One child died and two children withdrew before day 15, so day 15 endpoint data were available for 122 children for most endpoints, and 118 for the primary endpoint; all these children completed their allocated intervention and standard care and are included in the per protocol analysis. A further 2 children died and one withdrew between days 15 and 28. SAM, severe acute malnutrition. Hb, haemoglobin concentration.

Primary endpoint

The median day 15 composite faecal inflammatory score was lower in all treatment groups compared with the standard care group (Table  3 ). Using ANCOVA, with pre-specified P-value threshold of 0.10 and following adjustment for seven key covariates, pairwise comparisons showed the composite score in the teduglutide group was significantly lower than standard care (mean difference −0.89, 90%CI −1.69, −0.10, P = 0.07) (Table  4 ). Results were also stratified by site (Supplementary Table  S1 ). In Harare both teduglutide and budesonide reduced the primary outcome compared to standard care, by 2.1 (90%CI 0.7, 3.5; P = 0.02) and 1.4 (0.06, 2.7; P = 0.09) respectively, but the effect was not significant in Lusaka (Supplementary Table  S1 ). In the whole group, the ANCOVA model demonstrated no significant effect of sex, HIV infection, oedema, diarrhoea, or baseline weight-for-length Z-score (WLZ) on the faecal inflammatory score.

Secondary endpoints

Amongst secondary biomarker endpoints (Table  4 ), compared with standard care, budesonide reduced plasma CRP (mean reduction 5.2 mg/L; 90%CI 0.8, 9.6; P = 0.05) and sCD163 (mean reduction 405 ng/mL; 90%CI 73, 738; P = 0.05) while colostrum and N-acetyl glucosamine had effects only on CRP (reductions 5.9 mg/L; 90%CI 1.4, 10.3; P = 0.03, and 4.8 mg/L; 90%CI 0.5, 9.2; P = 0.07, respectively). There were notable site-specific differences in CRP, sCD14, sCD163 and iFABP (Supplementary Table  S1 ). Transformation of secondary endpoints to approximate a normal distribution did not change the effects observed, except for colostrum, for which an effect was found on GLP-2 only when the values were transformed (Supplementary Table  S2 ). N-acetyl glucosamine reduced mean days with diarrhoea by 89% (Table  4 ; ratio of NAG to standard care 0.11; 90%CI 0.04, 0.33; P = 0.001). None of the interventions affected days with fever or oedema, or change in weight, height, or MUAC.

Assessment of endoscopic biopsies

Of 62 children enrolled in Lusaka, endoscopy was carried out on 25 (5 colostrum, 4 N-acetyl glucosamine, 6 teduglutide, 5 budesonide, 5 Standard Care). Of the remainder, 9 children were unsuitable for anaesthesia (mainly upper respiratory infections), one child had no INR result available, 3 children missed endoscopy because no anaesthetist was available; and 24 children missed endoscopy because of instrument breakdowns. Representative specimens are shown in Fig.  2 . Morphometry in post-treatment biopsies showed significant differences in crypt depth (Fig.  3 ; P = 0.02 by Kruskal-Wallis test; by Dunn’s test P = 0.01 for colostrum and P = 0.049 for teduglutide compared to standard care).

Biopsies are from children treated with a colostrum, b N-acetyl glucosamine, c teduglutide, d budesonide, and e standard care. Morphometric analysis is shown in panel f . Scale bars show 200 μm. These biopsies were selected from 22 biopsies from 25 children: 6 in the teduglutide group, 5 in the colostrum group, 5 in the budesonide group, 5 in the standard care group, and 4 in the N-acetyl glucosamine group. Individual data from morphometric analysis are shown in Fig.  3 .

Measurements of villus height (VH) and crypt depth (CD) in 22 biopsies with satisfactory orientation obtained from children completing 14 days of treatment. a  villus height ( P = 0.84 by Kruskal-Wallis test across all groups). b crypt depth ( P = 0.02 by Kruskal-Wallis test; by Dunn’s test P = 0.01 for colostrum and P = 0.049 for teduglutide). c, epithelial surface area.  NAG, N-acetyl glucosamine. Source Data are provided as a Source Data File (Dataset 1).

Adverse events

A total of 102 clinical adverse events were reported (10 SAEs, 92 non-serious AEs) which did not differ significantly by treatment arm (Supplementary Table  S3 ). No AEs or SAEs were adjudicated as related to the investigative products. There were no adverse events related to endoscopy, and no Adverse Events of Special Interest. Laboratory AEs did not differ by treatment allocation (Supplementary Table  S4 ). Of the 10 SAEs observed, 3 (2% of the whole cohort) were deaths and 7 (6%) were prolonged hospitalisations or re-admission (Supplementary Table  S3 ). Of the three deaths, only one (on day 3) occurred before day 15; the other two occurred on days 20 and 23. Deaths were attributed to fever with diarrhoea ( n = 1; Standard Care arm), tuberculosis ( n = 1; Teduglutide arm), and aspiration pneumonia ( n = 1; Colostrum arm). Seven readmissions or prolonged hospitalisations occurred, due to fever ( n = 2), deterioration of oedema ( n = 1), vomiting ( n = 1), respiratory distress ( n = 1), and a burn from a hot water bottle whilst in hospital ( n = 1). One SAE was a readmission for observation at the day 28 visit after a protocol violation, due to receiving double the protocol dose of budesonide; there were no clinical or laboratory consequences. One child receiving teduglutide was readmitted for vomiting to exclude intestinal obstruction (an Adverse Event of Special Interest) but the illness resolved uneventfully and was adjudicated as unrelated to the investigative product. Laboratory abnormalities occurred in 702 instances, including baseline abnormalities, but did not differ significantly between arms and were all adjudicated as unrelated to the investigative products (Supplementary Table  S4 ).

Despite implementation of current treatment guidelines for complicated SAM, mortality remains unacceptably high, particularly in settings with high HIV burden. There is an urgent need for transformative approaches that modify underlying pathogenic pathways 39 . We identified intestinal mucosal damage as a promising target for intervention 27 . This multi-arm phase II clinical trial evaluated four potential new therapies to promote mucosal healing, which were each compared with standard care. Teduglutide showed benefit on the primary endpoint, a composite of three faecal inflammatory markers widely used to define enteropathy in malnourished children. Teduglutide is a GLP-2 agonist widely used in intestinal failure, a clinical syndrome with features in common with the more severe cases of malnutrition enteropathy. Our findings suggest that GLP-2 agonism similarly enhances mucosal healing in children with SAM. No other intervention significantly differed from standard care for the primary outcome. However, budesonide, colostrum and N-acetyl glucosamine reduced the systemic inflammatory marker CRP, which is potentially clinically important as CRP is a predictor of mortality 40 . The increase in crypt depth observed with teduglutide and bovine colostrum likely indicates enhanced mucosal regeneration, as these agents both showed some reductions in inflammatory biomarkers (faecal composite score for teduglutide, CRP for colostrum). N-acetyl glucosamine reduced diarrhoea, which is independently associated with mortality in SAM, potentially reflecting restoration of glycocalyx composition 32 and/or inhibition of enteropathogen colonisation 41 . Collectively, our data identify teduglutide as the leading candidate for future trials, but also suggest there may be benefits from the other treatments, which all have distinct mechanisms of action. This trial highlights the importance of measuring multiple biomarkers, which capture different pathological domains of malnutrition enteropathy. A larger-scale trial of single or combined interventions with outcomes including mortality and readmission is now warranted.

Our hypothesis was that one or more trial interventions could aid mucosal healing, reducing enteropathy, inflammation, and microbial translocation. We selected three faecal biomarkers of mucosal damage and inflammation as a composite primary endpoint, on which we based our sample size calculations. However, we acknowledged a priori that our interpretation would draw on the full range of endpoints, since no single biomarker or composite score captures the complexity of pathogenesis linking mucosal damage to mortality 27 . The effects of trial treatments on different secondary endpoints suggests that each may benefit specific domains of gut dysfunction in SAM 37 .

The potential usefulness of teduglutide is limited by its expense and subcutaneous route of administration. However, many therapies introduced at high prices fall in cost once volume of use increases. By contrast budesonide, which reduced the inflammatory markers CRP and sCD163, is far less costly and easier to administer. Neither should be implemented as part of treatment for SAM without further trial evidence, but the TAME trial demonstrates that use is likely to be safe, and confirms mucosal healing as a promising strategy in severe malnutrition. Although colostrum did not affect the primary endpoint, it increased both plasma levels of GLP-2 (though only after log transformation) and crypt depth. Colostrum contains GLP-2 at around 5 ng/ml and supplementation with colostrum in juvenile pigs undergoing intestinal resection increased plasma GLP-2 levels 42 .

As expected, adverse events were common but serious adverse events uncommon, and there were no events considered related to trial medications. The safety of long-term teduglutide has been assessed in intestinal failure and considered acceptable 35 . No Adverse Events of Special Interest were observed. A trial of mesalazine, another anti-inflammatory drug used in IBD, suggested safety in children with SAM 43 , and our data extend these findings by showing that these medications are also safe. Mortality was low compared to our historical data 9 , 11 . This may be due to the enhanced medical and nursing care usually associated with conduct of a clinical trial, but probably also relates to our caution in focusing on inclusion of clinically stable children. Given our experience of high early mortality in children with complicated SAM, we adopted this strategy to reduce the likelihood of serious adverse events in this phase II trial. In future it would appear desirable to bring forward treatments to the day of admission, when they might be of greatest potential benefit. Treatment duration was 14 days, representing the period of greatest mortality risk in hospitalised children; however, we and others have reported that unacceptably high mortality continues for 48 weeks following hospital discharge 2 , 11 , 13 . It is therefore possible that longer duration of therapies for mucosal healing could be of benefit.

We recognise several limitations. Due to restrictions on recruitment and difficulties transporting medicines and reagents during the COVID-19 pandemic, our enrolment was reduced. However, withdrawal and mortality were much lower than anticipated and we could therefore retain adequate power for the primary endpoint with a smaller sample size. Because this trial was conducted in hospital, adherence to medication was very high, with 98% of children completing intended therapy. This may be unattainable in real-world settings, but overall this trial demonstrated proof of concept for the therapies chosen. Our results were consistent across endpoint domains and biologically plausible for known mechanism of action of each agent, increasing confidence in our findings. Due to the short intervention duration, we saw limited impact on clinical outcomes such as growth. However, future trials powered for clinically important outcomes could explore the optimal timing, dosage and duration of intervention. There are also some challenges in interpreting the biomarkers used in this trial. Faecal biomarkers are subject to dilutional considerations, especially when a significant proportion of the trial participants have diarrhoea. It is also true that the biology of these biomarkers is not fully established. Markers such as myeloperoxidase and neopterin are elaborated by leukocytes, and reflect intestinal mucosal inflammation, but intestinal permeability and trafficking of leukocytes to the gut can be altered in the presence of systemic infections 44 , 45 . Neopterin is synthesised in response to interferon-γ and generally reflects Th1-mediated inflammation. α1-antitrypsin is usually considered a biomarker of protein loss into the gut, but transcriptomic data reveal that it is expressed in the mucosa 26 , 46 . These considerations need to be taken into account in future work. We have no ready explanation for the heterogeneity between study sites. We have previously noted mortality differences between Zambia and Zimbabwe, and there are minor differences in protocol implementation (such as when children are ready for discharge) which might explain some of these effects. This heterogeneity also needs to be taken into account in future work as it underlines the value of performing trials in different countries.

Our findings demonstrate a biologically plausible new treatment paradigm for children with complicated SAM. Intestinal damage is ubiquitous in children with SAM, driving systemic inflammation, contributing to stunting and developmental impairment and increasing mortality. No interventions for malnutrition enteropathy are currently available. We have shown that a short course of therapy added to standard care, aimed at restoring mucosal integrity, can ameliorate underlying pathogenic pathways. Further trials should evaluate these interventions for their effects on mortality, clinical recovery and long-term nutritional restoration to improve the outcomes of children with complicated SAM. Combinations of interventions would be interesting to evaluate in future trials since their distinct mechanisms of action and potential to target multiple domains of malnutrition enteropathy concurrently, may plausibly lead to greater mucosal healing and clinical recovery through synergistic effects.

Ethics approval was obtained from the University of Zambia Biomedical Research Ethics Committee (006-09-17), the National Health Research Committee of Zambia, the Zambia Medicines Regulatory Authority (CT 082/18), the Joint Research Ethics Committee of Harare Central Hospital (JREC/66/19), the Medicines Control Authority of Zimbabwe (CT/176/2019), and the Medical Research Council of Zimbabwe (MRCZ/A/2458). The trial was conducted in accordance with the principles of the Declaration of Helsinki. The trial was monitored by a Data Safety and Monitoring Board. The trial was registered at www.clinicaltrials.gov (NCT03716115) and first posted on 23 rd October 2018, prior to patient enrolment, and the protocol was published 27 . A CONSORT checklist containing information reporting a randomized controlled trial has been included in Supplementary Note  1 . The trial protocol and statistical analysis plan are included in Supplementary Notes  2 and 3 .

Recruitment, inclusion and exclusion criteria

Children were recruited from 4 th May 2020 to 27 th April 2021, and the trial closed on 25 th May 2021 after the completion of the period of follow up of the last child. Potentially eligible children were identified by the study nursing teams, and written permission to screen was obtained from caregivers. A detailed information sheet in local languages was discussed with caregivers prior to seeking consent. Weight, length and mid-upper arm circumference (MUAC) were measured three times, and eligibility ascertained. Children of either sex were eligible for inclusion if aged 6-59 months and hospitalised with SAM (using WHO definitions: WLZ <−3, and/or MUAC <11.5cm, and/or bilateral pedal oedema), with written, informed consent from the primary caregiver. Children were excluded if unstable (shocked, hypothermic, hypoglycaemic, impaired consciousness), under 5kg body weight, had conditions impairing feeding, haemoglobin below 6 g/dl, or if their caregiver would not consent to child HIV testing or to remaining in hospital throughout the treatment course. Additional exclusion criteria were contraindications to any treatment or other factors that might prejudice study completion or analysis. No payments were made for participation, but transport refunds were made on discharge and on review to permit return home using public transport.

Trial procedures

Children were randomised equally to each of the four interventions, or standard care. Trial identification numbers were allocated sequentially in each site, with randomisation carried out by opening a sealed envelope bearing the corresponding number. The randomisation sequence was prepared by the trial statistician (KVB), stratified by study site, in random permuted blocks of variable size. Interventions began as soon as possible after baseline samples were collected. Children were managed by a team of nurses providing 24-hour cover, ensuring all treatments were administered and adverse events recorded. Study doctors (KC, GT) reported clinical progress daily using a standardised form. All treatment courses were 14 days. Samples for endpoint analysis were collected on day 15 (permissible window 15-19); children were then discharged if ready and reviewed on day 28 (window 28-42). Blood samples for safety monitoring (full blood count, renal and liver function, phosphate) were collected at baseline, 5 and 15 days post-randomisation. We enrolled children with predominantly oedematous SAM, and a high prevalence of HIV infection, and were not powered to evaluate effects in different subgroups of children. Tuberculosis was diagnosed clinically, and specifically searched for in any child with respiratory symptoms, using microscopy and culture of nasogastric aspirates whenever possible, chest X-ray and urine lipoarabinomannan according to clinical protocols.

Investigational products

Bovine colostrum (supplied by Colostrum UK) and N-acetyl glucosamine (Blackburn Distributions, UK) are nutraceuticals, regarded as safe and not licensed as medicines. They were provided as powder and encapsulated to ensure accurate dosing (colostrum 1.5 g 8-hourly throughout; N-acetyl glucosamine 300 mg 8-hourly days 1–7, 600 mg 8-hourly days 8–14). The dose of colostrum (4.5 g/day orally or via nasogastric tube) was chosen to be similar to those used in other published studies involving children. Ismail et al. treated premature infants with a dose of approx. 0.5–1 g colostrum/kg/day to examine gut immune priming 47 , and Barakat & Omar used 3 g/day of colostrum for children aged 6 months to 2 years suffering from acute diarrhoea 48 . The dose of N-acetyl glucosamine was based on our previous study in paediatric Crohn’s disease, where daily dosage of 6 grams augmented expression of epithelial glycosaminoglycans without evidence of adverse effects 33 . Intravenous doses of up to 100 mg have been tolerated in adults without adverse effects 33 and breastfed newborns consume 650–1500 mg of n-acetyl glucosamine per litre of human breast milk from well-nourished mothers 49 . Budesonide 0.5 mg and 1 mg respules (Alliance Healthcare, UK) are designed for nebuliser therapy but used off-licence orally for gastrointestinal therapy. These were opened on the ward and administered orally or by nasogastric tube. Dosage was 1 mg 8-hourly during days 1–7, 1 mg 12-hourly during days 8–11, and then 0.5 mg 12-hourly during days 12–14. The dosage was derived from studies in paediatric Crohn’s disease, where 9 mg enteric-coated daily budesonide proved equally efficacious to 40 mg prednisolone but with substantially reduced adverse effects 50 . Teduglutide (Revestive, Takeda) is licensed widely for intestinal failure but never previously evaluated in SAM. It was provided as 1.25 mg vials which were stored at 4–8 °C, and given by subcutaneous injection (0.05 mg/kg daily, based on weight measured on days 1 and 8). Prior stability testing carried out by Takeda confirmed that the opened vial is stable at 4–8 °C for 24 hours so each vial provided two doses, drawn 24 hours apart. Site rotation was marked on a map of anatomical sites as recommended by the manufacturer. The selected dose for teduglutide (0.05 mg/kg) was found to be the most effective dose in a phase 3, 12-week paediatric trial when compared to two lower doses of 0.025 mg/kg and 0.0125 mg/kg 51 . It is the dose currently approved by the FDA in the US and the EMA in Europe.

Evaluation of endpoints

The primary endpoint was the day 15 composite faecal inflammatory biomarker score, comprising myeloperoxidase, neopterin and α 1 -antitrypsin 27 . Secondary endpoints at day 15 were: changes in anthropometry; plasma biomarkers of enteropathy, microbial translocation and systemic inflammation (iFABP, LBP, CRP, sCD14, CD163, IGFBP-3, and GLP-2); days with diarrhoea, fever, and oedema; and adverse events. For children who were potty trained, stool was collected using a clean pot and then the required amount was transferred to a sterile stool container using a scoop. For those who are were not potty trained, diapers were used. The diapers were put inside out so that the plastic layer was next to the skin. A scoop was used to place a sample in a sterile stool container. The collected sample was then put in a cooler box with ice packs immediately. A sample transmittal form was used to keep track of the transit time from point of collection to receipt in the laboratory. Nursing staff stayed in communication with lab staff to ensure rapid delivery of samples to the laboratory.

Biomarkers were assayed by ELISA (Supplementary Table  S5 ) by laboratory scientists (KZ, KM, EB) blinded to study arm, and re-calculated independently (by RN and JS) from raw data on harmonised Gen5 software (Biotek/Agilent, Santa Clara, CA). Serum albumin and lipopolysaccharide, though pre-specified as endpoints, were not included due to failing quality control checks leading to low concordance between sites. IGF-1 values were very close to zero; as insufficient plasma was available for re-testing these data have been omitted. Lactulose/rhamnose urinary excretion tests were only performed on children undergoing endoscopy and only 14 valid data pairs were obtained; these data are therefore not shown.

A subgroup of children in Lusaka additionally underwent endoscopy for duodenal biopsy between days 15 and 19; only the Lusaka site was selected for this due to its considerable experience in paediatric endoscopy over many years. Except for two periods when endoscopy instruments required repair, children were selected sequentially, provided there were no haematological or anaesthetic contraindications. Sedation was administered by an anaesthetist (HS or MZ) and biopsies were collected from the second part of the duodenum using a Pentax 2490i paediatric gastroscope. Biopsies were orientated under a dissecting microscope and fixed before processing into paraffin blocks, sectioning and staining. Slides were scanned at 20x magnification on an Olympus VS-120 scanning microscope and blinded morphometry was performed by a single observer (CM, confirmed by PK) on all villus and crypt units identifiable in well-orientated parts of sections of each biopsy. The criterion used for suitable orientation was that crypts should be visualised throughout their length (see Fig.  2 , and reference 22 ), and then the boundary between crypt and villus compartments delineated. Crypt depth was measured in micrometres (μm) from this boundary to the furthest point of the base of the crypt, where the basement membrane would be expected. Villus height was measured in μm from the boundary to the villus tip in a straight line. Epithelial surface area was measured as the perimeter of the villi where muscularis mucosae could be measured, and expressed per micrometre of muscularis mucosae. Any portions of these sections where crypts were not visualised along their entire length were deemed poorly orientated and not used for morphometry. The median number of villi measured was 6 (IQR 4-9; range 3-13).

Adverse events between enrolment and day 15 or day 28 were reported in real time and reviewed for seriousness, severity, relatedness and expectedness; all serious adverse events were reported to the Sponsor (Queen Mary University of London), ethics committees and national trial regulators. Severity was categorised as mild, moderate or severe using the DAIDS classification ( https://rsc.niaid.nih.gov/clinical-research-sites/daids-adverse-event-grading-tables ), and all AEs reported monthly to the Data Monitoring and Ethics Committee (DMEC). Three Adverse Events of Special Interest (AESIs) were specifically sought: intestinal obstruction or volume overload for teduglutide, and osmotic diarrhoea for colostrum and N-acetyl glucosamine. Haematology and biochemistry results at baseline, days 5 and 15 were graded using DAIDS tables.

Sample size

The planned sample size was 225 children (45 in each arm), based on the composite biomarker score 27 . Enrolment was slowed by COVID-19, but trial losses were much lower than anticipated (3% observed versus 20% anticipated). The Trial Steering Committee and DMEC therefore reviewed the sample size in January 2021 once 82 children had been enrolled. The decision was made to reduce the sample size, based on a Cohen’s d effect size of 0.3, with 80% power and 90% confidence, and conservative correlation between baseline and follow-up estimates of 0.5, requiring 23 per group across 5 groups to analyse the primary outcome by ANCOVA. Allowing for 5% losses, the sample size of 115 was rounded up to 125 participants in total (25 per group).

Statistical analysis

Per protocol analyses were pre-specified 27 , and all hypothesis testing was 2-sided. Statistical analysis was performed in Stata 17 (Stata corp, College Station, Texas). Analysis of primary and secondary endpoints was based on comparison against standard care. ANCOVA was used to model final endpoint values, with adjustment for core baseline value, sex, baseline presence of oedema, HIV status, baseline diarrhoea, baseline WLZ score, and study site. Mortality was low (3 deaths) so could not be analysed statistically. Covariates chosen were pre-specified to take into account important elements of pathophysiology (worse outcome in HIV infection and oedematous malnutrition and in children with diarrhoea 52 ) and to allow for possible differences between the two countries. For some secondary endpoints negative binomial models were constructed which used a smaller set of adjustment variables (sex & HIV) due to model constraints (Table  2 ). Anthropometric measurements were calculated as change from baseline. The endoscopy subset was analysed separately, comparing post-treatment morphometric measurements by Kruskal-Wallis test, followed by Dunn’s test. Treatment effects were deemed statistically significant if the P value was <0.1 when compared to the control arm, as pre-specified. No adjustments of the false-positive (type I) error rate were planned, in line with the general consensus that adjustment for type I error rate is not required in exploratory multi-arm multi-stage trials in Phase II within the treatment development framework 27 , 53 .

Reporting summary

Further information on research design is available in the  Nature Portfolio Reporting Summary linked to this article.

Data availability

Data supporting the findings of this study are available in the article and in its Supplementary Information. Source data are provided as a Source Data file for Fig.  3 , and have also been deposited in Figshare under accession code https://doi.org/10.6084/m9.figshare.24442699 . The data uploaded to figshare include deidentified individual participant data, trial protocol, and statistical analysis plan.

Bhutta, Z. A. et al. Severe childhood malnutrition. Nat. Rev. Dis. Prim. 3 , 17067 (2017).

Article   PubMed   Google Scholar  

Childhood Acute Illness and Nutrition (CHAIN) Network. Childhood mortality during and after acute illness in Africa and south Asia: a prospective cohort study. Lancet Glob. Health 10 , e673–e684 (2022).

Article   Google Scholar  

WHO Guideline: Updates on the management of severe acute malnutrition in infants and children. Geneva: World Health Organization; 2013.

Lenters, L. M., Wazny, K., Webb, P., Ahmed, T. & Bhutta, Z. A. Treatment of severe and moderate acute malnutrition in low- and middle-income settings: a systematic review, meta-analysis and Delphi process. BMC Public Health 13 , S23 (2013).

Article   PubMed Central   PubMed   Google Scholar  

Trehan, I. et al. Antibiotics as part of the management of severe acute malnutrition. N. Engl. J. Med. 368 , 425–435 (2013).

Article   CAS   PubMed Central   PubMed   Google Scholar  

Isanaka, S. et al. Routine Amoxicillin for uncomplicated severe acute malnutrition in children. N. Engl. J. Med. 374 , 444–453 (2016).

Article   CAS   PubMed   Google Scholar  

Tickell, K. D. & Denno, D. M. Inpatient management of children with severe acute malnutrition: a review of WHO guidelines. Bull. World Health Org. 94 , 642–651 (2016).

Nalwanga, D. et al. Mortality among children under five years admitted for routine care of severe acute malnutrition: a prospective cohort study from Kampala, Uganda. BMC Pediatr. 20 , 182 (2020).

Amadi, B. C. et al. Intestinal and systemic infection, HIV and mortality in Zambian children with persistent diarrhoea and malnutrition. J. Ped Gastroenterol. Nutr. 32 , 550–554 (2001).

CAS   Google Scholar  

Kerac, M. et al. Follow-up of post-discharge growth and mortality after treatment for severe acute malnutrition (FuSAM study): a prospective cohort study. PLoS One 9 , e96030 (2014).

Article   ADS   PubMed Central   PubMed   Google Scholar  

Bwakura-Dangarembizi, M. et al. Risk factors for post-discharge mortality following hospitalization for severe acute malnutrition in Zimbabwe and Zambia. Am. J. Clin. Nutr. 113 , 665–674 (2021).

Girma, T. et al. Nutrition status and morbidity of Ethiopian children after recovery from severe acute malnutrition: Prospective matched cohort study. PLoS One 17 , e0264719 (2022).

Njunge, J. M. et al. Biomarkers of post-discharge mortality among children with complicated severe acute malnutrition. Sci. Rep. 9 , 5981 (2019).

Bhutta, Z. A., Guerrant, R. L. & Nelson, C. A. 3rd Neurodevelopment, nutrition, and inflammation: the evolving global child health landscape. Pediatrics 139 , S12–S22 (2017).

Kar, B. R., Rao, S. L. & Chandramouli, B. A. Cognitive development in children with chronic protein energy malnutrition. Behav. Brain Funct. 4 , 31 (2008).

Galler, J. R. et al. Neurodevelopmental effects of childhood malnutrition: A neuroimaging perspective. NeuroImage 231 , 117828 (2021).

Prendergast, A. S. J. & Kelly, P. Interactions between intestinal pathogens, enteropathy and malnutrition in developing countries. Curr. Opin. Infect. Dis. 29 , 229–236 (2016).

Brunser, O., Castillo, C. & Araya, M. Fine structure of the small intestinal mucosa in infantile marasmic malnutrition. Gastroenterology 70 , 495–507 (1976).

Shiner, M., Redmond, A. O. & Hansen, J. D. The jejunal mucosa in protein-energy malnutrition. A clinical, histological, and ultrastructural study. Exp. Mol. Pathol. 19 , 61–78 (1973).

Attia, S. et al. Mortality in children with complicated severe acute malnutrition is related to intestinal and systemic inflammation: an observational cohort study. Am. J. Clin. Nutr. 104 , 1441–1449 (2016).

Amadi, B. et al. Impaired barrier function and autoantibody generation in malnutrition enteropathy in Zambia. EBiomedicine 22 , 191–199 (2017).

Mulenga, C. et al. Epithelial abnormalities in the small intestine of Zambian children with stunting. Front Med (Gastroenterol.) 9 , 849677 (2022).

Liu, T.-C. et al. A novel histological index for evaluation of environmental enteric dysfunction identifies geographic-specific features of enteropathy among children with suboptimal growth. PLoS Negl. Trop. Dis. 14 , e0007975 (2020).

Campbell, D. I., Murch, S. H. & Elia, M. Chronic T cell-mediated enteropathy in rural West African children: relationship with nutritional status and small bowel function. Pediatr. Res. 54 , 306–311 (2003).

Brenchley, J. M. & Douek, D. C. Microbial translocation across the GI tract. Annu Rev. Immunol. 30 , 149–173 (2012).

Chama, M. et al. Transcriptomic analysis of enteropathy in Zambian children with severe acute malnutrition. EBioMedicine 45 , 456–463 (2019).

Kelly, P. et al. TAME trial: a multi-arm phase 2 randomised trial of four novel interventions for malnutrition enteropathy in Zambia and Zimbabwe; a study protocol. BMJ Open 9 , e027548 (2019).

Playford, R. J. & Weiser, M. J. Bovine Colostrum: Its constituents and uses. Nutrients 13 , 265 (2021).

Chandwe, K. & Kelly, P. Colostrum therapy for human gastrointestinal health and disease. Nutrients 13 , 1956 (2021).

Khan, Z. et al. Use of the ‘nutriceutical’, bovine colostrum, for the treatment of distal colitis: results from an initial study. Aliment Pharm. Ther. 16 , 1917–1922 (2002).

Article   CAS   Google Scholar  

Amadi, B. et al. Reduced production of sulfated glycosaminoglycans occurs in Zambian children with kwashiorkor but not marasmus. Am. J. Clin. Nutr. 89 , 592–600 (2009).

Bode, L. et al. Heparan sulfate and syndecan-1 are essential in maintaining murine and human intestinal epithelial barrier function. J. Clin. Invest 118 , 229–238 (2008).

Salvatore, S. et al. A pilot study of N-acetyl glucosamine, a nutritional substrate for glycosaminoglycan synthesis, in paediatric chronic inflammatory bowel disease. Aliment Pharm. Ther. 14 , 1567–1579 (2000).

Kocoshis, S. A. et al. Safety and efficacy of Teduglutide in pediatric patients with intestinal failure due to short Bowel Syndrome: A 24-week, Phase III Study. JPEN J. Parenter. Enter. Nutr. 44 , 621–631 (2020).

Hill, S. et al. Safety findings in pediatric patients during long-term treatment with teduglutide for short-bowel syndrome-associated intestinal failure: pooled analysis of 4 clinical studies. JPEN J. Parenter. Enter. Nutr. 45 , 1456–1465 (2021).

Rezaie, A. et al. Budesonide for induction of remission in Crohn’s disease. Cochrane Database Syst. Rev. 2015 , CD000296 (2015).

PubMed Central   PubMed   Google Scholar  

Owino, V. et al. Environmental Enteric dysfunction and growth failure/stunting in global child health. Pediatrics 138 , e20160641 (2016).

Trehan, I., Kelly, P., Shaikh, N. & Manary, M. J. New insights into environmental enteric dysfunction. Arch. Dis. Child 101 , 741–744 (2016).

Picot, J. et al. The effectiveness of interventions to treat severe acute malnutrition in young children: a systematic review. Health Tech. Assess. 16 , 19 (2012).

Article   MathSciNet   Google Scholar  

Wen, B. et al. Systemic inflammation and metabolic disturbances underlie inpatient mortality among ill children with severe malnutrition. Sci. Adv. 8 , eabj6779 (2022).

Article   ADS   CAS   PubMed Central   PubMed   Google Scholar  

Sicard, J. F. et al. N-acetyl-glucosamine influences the biofilm formation of Escherichia coli . Gut Pathog. 10 , 26 (2018).

Paris, M. C. et al. Plasma GLP-2 levels and intestinal markers in the juvenile pig during intestinal adaptation: effects of different diet regimens. Dig. Dis. Sci. 49 , 1688–1695 (2004).

Jones, K. D. et al. Mesalazine in the initial management of severely acutely malnourished children with environmental enteric dysfunction: a pilot randomized controlled trial. BMC Med 12 , 133 (2014).

Wang, J. et al. Respiratory influenza virus infection induces intestinal immune injury via microbiota-mediated Th17 cell-dependent inflammation. J. Exp. Med. 211 , 2397–2410 (2014). Erratum in: J Exp Med. 2014 Dec 15;211(13):2683. Erratum in: J Exp Med 2014;211(12):2396-7.

Sturgeon, J. P., Bourke, C. D. & Prendergast, A. J. Children with noncritical infections have increased intestinal permeability, endotoxemia and altered innate immune responses. Pediatr. Infect. Dis. J. 38 , 741–748 (2019).

Kelly, P. et al. Gene expression profiles compared in environmental and malnutrition enteropathy in Zambian children and adults. EBioMedicine 70 , 103509 (2021).

Ismail, R. I. H. et al. Gut priming with bovine colostrum and T regulatory cells in preterm neonates: a randomized controlled trial. Pediatr. Res. 90 , 650–656 (2021).

Barakat, S. H., Meheissen, M. A., Omar, O. M. & Elbana, D. A. Bovine colostrum in the treatment of acute diarrhea in children: a double-blinded randomized controlled trial. J. Trop. Pediatr. 66 , 46–55 (2020).

PubMed   Google Scholar  

Miller, J. B., Bull, S., Miller, J. & McVeagh, P. The oligosaccharide composition of human milk: temporal and individual variations in monosaccharide components. J. Pediatr. Gastroenterol. Nutr. 19 , 371–376 (1994).

Levine, A. et al. A comparison of budesonide and prednisone for the treatment of active pediatric Crohn disease. J. Pediatr. Gastroenterol. Nutr. 36 , 248–252 (2003).

CAS   PubMed   Google Scholar  

Carter, A. et al. Outcomes from a 12-week, open-label, multicenter clinical trial of teduglutide in pediatric short bowel syndrome. J. Pediatr. 181 , 102–111 (2017).

Karunaratne, R., Sturgeon, J. P., Patel, R. & Prendergast, A. J. Predictors of inpatient mortality among children hospitalized for severe acute malnutrition: a systematic review and meta-analysis. Am. J. Clin. Nutr. 112 , 1069–1079 (2020).

Wason, J. M., Stecher, L. & Mander, A. P. Correcting for multiple-testing in multi-arm trials: is it necessary and is it done? Trials 15 , 364 (2014).

Download references

Acknowledgements

We are grateful to the following nurses from the wards and endoscopy unit of UTH: Evelyn Nyendwa, Esther Chilala, Andreck Tembo, Lucy Macwani, Dalitso Tembo, Mary Phiri, Elaine Brittel Sikuyuba, Sophreen Mwaba, Gwendolyn Nayame, Joyce Sibwani, Rose Soko, Kashinga Maseko, and Mulima Mwiinga. We are grateful to the nursing team in Harare Central Hospital: Sarudzai Murumbi, Tariro Zure, and Lucia Manyatera. We also sincerely thank Mr Rizvan Batha of Barts Health Pharmacy for assistance with procurement of investigational products. We are very grateful to the Trial Steering committee (Professors Jay Berkley, Ian Sanderson, and James Wason) and Data Monitoring and Ethics Committee (Professor Jim Todd, Doctors Rose Kambarami, Veronica Mulenga, and Philip Ayieko). The TAME trial was sponsored by Queen Mary University of London, but the Sponsor played no role in study design, data collection, analysis, or manuscript writing. The trial was funded by a grant from the Medical Research Council (UK), number MR/P024033/1. AJP and JPS are funded by Wellcome (108065/Z/15/Z for AJP, and 220566/Z/20/Z for JPS). Takeda UK provided teduglutide at a discounted price.

The Medical Research Council (UK) funded the study. Takeda UK provided tedu-glutide at a discounted price.

Author information

These authors contributed equally: Kanta Chandwe, Mutsa Bwakura-Dangarembizi.

Authors and Affiliations

Tropical Gastroenterology & Nutrition group, University of Zambia School of Medicine, Nationalist Road, Lusaka, Zambia

Kanta Chandwe, Beatrice Amadi, Deophine Ngosa, Nivea Chulu, Kanekwa Zyambo, Chola Mulenga, Ellen Besa, Bwalya Simunyola, Lydia Kazhila, Miyoba Chipunza, Victor Mudenda, Kelley VanBuskirk & Paul Kelly

Zvitambo Institute for Maternal and Child Health Research, McLaughlin Avenue, Meyrick Park, Harare, Zimbabwe

Mutsa Bwakura-Dangarembizi, Gertrude Tawodzera, Anesu Dzikiti, Robert Makuyana, Kuda Mutasa, Jonathan P. Sturgeon, Shepherd Mudzingwa, Batsirai Mutasa, Virginia Sauramba, Lisa Langhaug & Andrew J. Prendergast

Faculty of Medicine and Health Sciences, University of Zimbabwe, Parirenyatwa Hospital, Harare, Zimbabwe

Mutsa Bwakura-Dangarembizi

Blizard Institute, Queen Mary University of London, Newark Street, London, UK

Jonathan P. Sturgeon, Andrew J. Prendergast & Paul Kelly

Department of Anaesthesia, University of Zambia School of Medicine, Nationalist Road, Lusaka, Zambia

Masuzyo Zyambo & Hazel Sonkwe

Warwick University Medical School, Coventry, UK

Simon H. Murch

Great Ormond Street Hospital, London, UK

University of West London, Ealing, London, UK

Raymond J. Playford

University College Cork, College Road, Cork, Ireland

You can also search for this author in PubMed   Google Scholar

Contributions

KC, MBD, BA, SHM, RJP, SH, AJP and PK initiated and designed the trial. KC, GT, DN, AD, NC, RM, LK, BM and LL designed the trial instruments and data collection procedures. KC, MBD, BA, GT, DN, AD, NC, and RM carried out the daily clinical care and data collection. KZ, KM, CM, EB and VM were responsible for laboratory operating procedures, data acquisition and processing. LK, BM and LL undertook data entry and cleaning. MC, VS and LL undertook monitoring and quality control of the trial. BS and SM designed and implemented the pharmacy and pharmacovigilance procedures. MZ and HS undertook anaesthetic procedures and ensured the safety of children undergoing endoscopy. Analysis was carried out by KVB, JPS, LL, AJP and PK; AJP and PK wrote the initial draft which was revised by all authors.

Corresponding author

Correspondence to Paul Kelly .

Ethics declarations

Competing interests.

RJP was previously an external consultant to Colostrum UK which provided the bovine colostrum used in these studies. RJP has also been an external consultant to Sterling Technology (USA) and an employee of Pantheryx Inc (USA) who produce and distribute bovine colostrum. There was no bovine colostrum company involvement in the production of this article or editing of its content. SH has had funding for teduglutide studies and lectured and participated in advisory boards on behalf of Takeda. The remaining authors declare no competing interests.

Peer review

Peer review information.

Nature Communications thanks Indi Trehan and the other anonymous, reviewers for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information, peer review file, reporting summary, source data, source data, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Chandwe, K., Bwakura-Dangarembizi, M., Amadi, B. et al. Malnutrition enteropathy in Zambian and Zimbabwean children with severe acute malnutrition: A multi-arm randomized phase II trial. Nat Commun 15 , 2910 (2024). https://doi.org/10.1038/s41467-024-45528-0

Download citation

Received : 28 April 2023

Accepted : 26 January 2024

Published : 17 April 2024

DOI : https://doi.org/10.1038/s41467-024-45528-0

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here .

Loading metrics

Open Access

Peer-reviewed

Research Article

Modelling chronic malnutrition in Zambia: A Bayesian distributional regression approach

Contributed equally to this work with: Given Moonga, Johannes Seiler

Roles Conceptualization, Formal analysis, Methodology, Visualization, Writing – original draft, Writing – review & editing

* E-mail: [email protected]

Affiliations Center for International Health, Ludwig Maximilian University of Munich, Munich, Germany, Department of Public Health, Health Services Research and Health Technology Assessment, UMIT—University for Health Sciences, Medical Informatics and Technology, Hall in Tirol, Austria, Department of Epidemiology and Biostatistics, University of Zambia, Lusaka, Zambia

Roles Conceptualization, Supervision, Writing – review & editing

¶ ‡ These authors also contributed equally to this work.

Affiliations Department of Public Health, Health Services Research and Health Technology Assessment, UMIT—University for Health Sciences, Medical Informatics and Technology, Hall in Tirol, Austria, Institute and Outpatient Clinic for Occupational, Social and Environmental Medicine, Clinical Centre of the Ludwig Maximilian University of Munich, Munich, Germany

Roles Methodology, Supervision, Writing – review & editing

Affiliation Institute for medical Information Processing, Biometry, and Epidemiology, Ludwig Maximilian University of Munich, Munich, Germany

Roles Data curation, Supervision, Writing – review & editing

Affiliation Department of Humanities, Social and Political Sciences, ETH Zurich, Zurich, Switzerland

Roles Supervision

Affiliation Department of Epidemiology and Biostatistics, University of Zambia, Lusaka, Zambia

Roles Supervision, Writing – review & editing

Affiliation Institute and Outpatient Clinic for Occupational, Social and Environmental Medicine, Clinical Centre of the Ludwig Maximilian University of Munich, Munich, Germany

Affiliation Department of Public Health, Health Services Research and Health Technology Assessment, UMIT—University for Health Sciences, Medical Informatics and Technology, Hall in Tirol, Austria

Affiliation School of Veterinary Medicine, University of Zambia, Lusaka, Zambia

Roles Conceptualization, Data curation, Formal analysis, Methodology, Visualization, Writing – original draft, Writing – review & editing

Affiliations Department of Public Health, Health Services Research and Health Technology Assessment, UMIT—University for Health Sciences, Medical Informatics and Technology, Hall in Tirol, Austria, Department of Statistics, University of Innsbruck, Innsbruck, Austria

• Given Moonga, 
• Stephan Böse-O’Reilly, 
• Ursula Berger, 
• Kenneth Harttgen, 
• Charles Michelo, 
• Dennis Nowak, 
• Uwe Siebert, 
• John Yabe, 
• Johannes Seiler

• Published: August 4, 2021
• https://doi.org/10.1371/journal.pone.0255073
• Reader Comments

The burden of child under-nutrition still remains a global challenge, with greater severity being faced by low- and middle-income countries, despite the strategies in the Sustainable Development Goals (SDGs). Globally, malnutrition is the one of the most important risk factors associated with illness and death, affecting hundreds of millions of pregnant women and young children. Sub-Saharan Africa is one of the regions in the world struggling with the burden of chronic malnutrition. The 2018 Zambia Demographic and Health Survey (ZDHS) report estimated that 35% of the children under five years of age are stunted. The objective of this study was to analyse the distribution, and associated factors of stunting in Zambia.

We analysed the relationships between socio-economic, and remote sensed characteristics and anthropometric outcomes in under five children, using Bayesian distributional regression. Georeferenced data was available for 25,852 children from two waves of the ZDHS, 31% observation were from the 2007 and 69% were from the 2013/14. We assessed the linear, non-linear and spatial effects of covariates on the height-for-age z-score.

Stunting decreased between 2007 and 2013/14 from a mean z-score of 1.59 (credible interval (CI): -1.63; -1.55) to -1.47 (CI: -1.49; -1.44). We found a strong non-linear relationship for the education of the mother and the wealth of the household on the height-for-age z-score. Moreover, increasing levels of maternal education above the eighth grade were associated with a reduced variation of stunting. Our study finds that remote sensed covariates alone explain little of the variation of the height-for-age z-score, which highlights the importance to collect socio-economic characteristics, and to control for socio-economic characteristics of the individual and the household.

Conclusions

While stunting still remains unacceptably high in Zambia with remarkable regional inequalities, the decline is lagging behind goal two of the SDGs. This emphasises the need for policies that help to reduce the share of chronic malnourished children within Zambia.

Citation: Moonga G, Böse-O’Reilly S, Berger U, Harttgen K, Michelo C, Nowak D, et al. (2021) Modelling chronic malnutrition in Zambia: A Bayesian distributional regression approach. PLoS ONE 16(8): e0255073. https://doi.org/10.1371/journal.pone.0255073

Editor: Tzai-Hung Wen, National Taiwan University, TAIWAN

Received: February 23, 2020; Accepted: July 10, 2021; Published: August 4, 2021

Copyright: © 2021 Moonga et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: Data are third party data available from the following sites: https://crudata.uea.ac.uk/cru/data/drought/ https://malariaatlas.org/ https://ngdc.noaa.gov/eog/dmsp/downloadV4composites.html https://sedac.ciesin.columbia.edu/data/set/gpw-v4-population-density-rev11/data-download https://dhsprogram.com/data/ .

Funding: Johannes Seiler acknowledges financial support from the University of Innsbruck through a postdoctoral scholarship. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: No authers have competing interests.

Introduction

The burden of child malnutrition still remains a global challenge, with greater severity being faced by low-and middle-income countries [ 1 – 3 ]. Globally, malnutrition is amongst the most important risk factors associated with illness and death, affecting hundreds of millions of pregnant women and young children [ 3 – 6 ]. Stunting in early childhood is strongly associated with numerous short-term and long-term consequences, including increased childhood morbidity and mortality, delayed growth and motor development and long-term educational and economic consequences later in life [ 7 ]. Undernourishment causes children to start life at mentally suboptimal levels [ 8 ].

Assessment of childhood malnutrition commonly relies on standard anthropometric measures for insufficient height-for-age (stunting) indicating chronic undernutrition, insufficient weight-for-height (wasting), indicating acute undernutrition; and insufficient weight-for-age (underweight), an indicator commonly used to asses, both, chronic, and acute undernutrition [ 9 , 10 ].

Anthropometric measurements are practical techniques for assessing children’s growth patterns during the first years of life. The measurements also provide useful insights into the nutrition and health situation of entire population groups. Anthropometric indicators are less accurate than clinical and biochemical techniques in assessing individual nutritional status. However in resources limited settings, the measurements are a useful screening tool to identify individuals at risk of undernutrition, who can later be referred to subsequent possible confirmatory investigation [ 11 ].

It is estimated that globally 52 million children under-five years of age are wasted, 17 million are severely wasted and 155 million are stunted. Around 45% of deaths among children under-five years of age, most of which occur in the sub-Saharan Africa are linked to undernutrition [ 3 , 9 ]. It is also estimated that four out of ten children under the age of five in Zambia are stunted [ 12 ]. This paper will therefore focus on childhood stunting in Zambia.

Global prevalence of stunting in children younger than five years declined during the past two decades, but still remain unacceptably high in South Asia and sub-Saharan Africa regions [ 5 ]. If current trends remain unchecked, projections indicate that 127 million children under five years of age will be stunted in 2025 [ 1 ]. There is therefore need to heighten various interventions in these affected region and to investigate possible area specific determinants of stunting.

There are already fairly well documented perspectives on determinants of malnutrition. The treatise on these determinants mainly relies on the United Nations Children’s Fund (UNICEF) conceptual framework on malnutrition which has evolved over time as more knowledge and evidence on the causes, consequences and impacts of undernutrition is generated. The framework distinguishes between immediate, intermediate and underlying determinants of malnutrition [ 5 , 13 – 15 ].

The immediate causes of undernutrition include inadequate dietary intake and disease, while the underlying causes could include household food insecurity, inadequate care and feeding practices for children, unhealthy household and surrounding environments, and inaccessible and often inadequate health care. Basic causes of poor nutrition encompasses the societal structures and processes that neglect human rights and perpetuate poverty, constraints faced by populations to essential resources [ 13 ].

Several studies done within sub-Saharan Africa investigated determinants such as the mother’s level of education, income levels and these factors have been linked to malnutrition [ 9 , 12 , 16 , 17 ]. The source of the drinking water, the wealth of the household, the area of residence, age of the child, the sex of the child, the breastfeeding duration, the age of the mother has also been investigated and were observed to be significant correlates of stunting [ 12 , 18 ]. Within Zambia, stunting was observed to be more likely among children of less educated mothers (45%) and those from the poorest households (47%) [ 19 ]. The determinates of malnutrition are related to each other and the differences and direction between these levels of determinism as indicated in the UNICEF framework are often not discrete but in reality related. As discussed by Kandala [ 17 ] for example, the mother’s level of education might be influencing child care practises- an intermediate determinant—and the resources available to the household—an underlying determinant.

Previous studies elsewhere have observed that stunting tends to show regional variation [ 4 , 9 , 16 ]. We see this trend in Zambia as well, where the decline of stunting has been only gradual and unacceptable, with higher prevalence in Northern province where 50% of the children being stunted, and stunting being less common in Lusaka, Copperbelt, and Western provinces where 36% of children are stunted [ 19 ]. We see this regional variation of stunting in Fig 1 which shows stunting in Zambia in the 2007 and 2013/14 waves of the Zambian Demographic and Health Surveys (ZDHS). The ZDHS is a national-wide survey which is representative at a sub-national level and contains information on trends in fertility, childhood mortality, use of family planning methods, and maternal and child health indicators including HIV and AIDS [ 19 ]. The figure shows the height-for-age z-score, with Western province better than Northern province for the 2007 wave. We see a slight difference in 2013/14 as stunting seemed to get worse in parts of the Western province.

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

The panel shows the average height-for-age z-score at district level for 2007 (left) and 2013/14 (right) of Zambia. Source : Demographic and Health Surveys (data) and Database of Global Administrative Areas (boundary information); calculation by authors. The shapefile used to create these maps is republished from [ 54 ] under CC BY license, with permission from Robert J. Hijmans, original copyright [2021].

https://doi.org/10.1371/journal.pone.0255073.g001

Much of the work done on the determinants of stunting in Zambia, have considered socio-economic characteristics and have assessed the linear effects of these determinants on the conditional mean [ 12 , 20 ], using models specifications such as; linear models, generalized linear models (GLMs) and generalized additive models (GAMs) [ 21 ]. These aproaches are useful and have the advantage of being easy to estimate and to interpret. However, they may risk model misspecification and draw inaccurate estimates, when heterogeneity is present, or when extreme values in the response are present and when a linear relationship is not plausible. In the analysis of certain outcomes, like stunting for example, the interest is not only in the conditional mean, but also in extreme values (height-for-age z-scores), or other parameters of the response. Quantile regression is one possibility to model beyond the conditional mean, with the interest to show variation of the outcome at a quantile level, without making any assumptions of the response distribution. This method for example has been applied in child malnutrition studies [ 16 ]. However, distributional regression offers advantage over quantile regression, as it provides the possibility to characterize the complete probabilistic distribution of the response in one joint model [ 21 , 22 ]. Moreover, distributional regression is more efficient, if prior knowledge on specific aspects of the response distribution is available, or can be estimated [ 23 ]. Furthermore, the characterization of the whole distribution of the response is more informative.

The study by Kandala [ 24 ] which focused on stunting in sub-Saharan countries found that there are distinct spatial patterns of malnutrition that are not explained by the socio-economic determinants or other well-known correlates alone [ 16 , 25 ]. As such, our study includes spatial covariates, since we aim at investigating spatial differences of stunting in Zambia at sub-district level while jointly analysing socio-economic and environmental characteristics.

The following three reasons make our study novel compared to previous work [ 18 ]. Firstly, we jointly analysed remote sensed data and socio-economic covariates at sub-national level. This is made possible due to availability of georeferenced data at the primary sampling unit a household pertains to in the recent two ZDHS datasets. The Demographic and Health Surveys rely in most cases on a two-stage survey, and the primary sampling units corresponds to the enumeration areas from the most recent completed census that has been selected. Georeferenced data is important as it generates more specific information which can facilitate targeted interventions. Secondly, we used Bayesian distributional regression which allows us to model all parameters of the underlying response distribution. Lastly, we used two waves of the demographic health surveys to control for spatio temporal interactions. Therefore, this study demonstrates small area variation in stunting in Zambia and analyse possible inequalities and deprivation at the sub-district level.

Data sources

Socio-economic and georeferenced covariates.

We used data from the 2007 and 2013/14 ZDHS. The ZDHS is a national-wide survey which is representative at a sub-national level and contains information on trends in fertility, childhood mortality, use of family planning methods, and maternal and child health indicators including HIV and AIDS. For these population health indicators, data is collected for women aged 15–49, men aged 15–59 and children below five years of age [ 19 ].

The ZDHS provide besides information on the district a household pertains to, also information about the geolocation of the primary sampling unit a household belongs to, and from which the data was collected. The location of the primary sampling unit is the spatial information used in the empirical analysis. During data processing, GPS coordinates are displaced to ensure that respondent confidentiality is maintained. The displacement is randomly applied so that rural points contain a minimum of 0 and a maximum of 5 km of positional error. Urban points contain a minimum of 0 and a maximum of 2 km of error. A further 1% of the rural sample points are offset a minimum of 0 and a maximum of 10 km [ 26 ].

Demographic Health Surveys have documented weakness for estimation of individual anthropometric measurements. Potential threats to high data quality may occur across various research stages, from survey design to data analysis. There is also often a substantial amount of missing or implausible anthropometric data across surveys [ 27 ].

Furthermore, there is caution over the use of stunting as an individual classifier in epidemiologic research or its interpretation as a clinically meaningful health outcome. Stunting should be used as originally designed to be from its original use as a population level indicator of community well-being [ 28 ], as it reflects past health and nutrition conditions; and an indication of socio-economic development of a country [ 1 ].

Despite the above highlighted limitations of DHS and anthropometric indicators, they remain useful national wide measurements that can be used to estimate child health. Moreover, in general anthropometric measures are a good indicator for planning as they can provide a lot of information to policy makers to answer, how, where and which type of intervention would be favourable in specific settings.

Socio-economic and spatial determinants

The effects of socio-economic factors, such as the education of the mother, household size, wealth of the household on the health status of children are well documented [ 12 , 29 ]. We calculated an index representing the wealth of the household based on the household’s assets using Principal Components Analysis (PCA) following Filmer and Pritchett, and Sahn [ 30 , 31 ]. Previous studies have shown that household wealth status was a predictor of childhood malnutrition. Children from poor households are more likely to be stunted than those from richer households [ 29 ].

In our analysis we investigated the impact of different socio-economic factors, which impact on height-for age Z-score has been discussed in literature. Table 1 gives an overview and the according references.

https://doi.org/10.1371/journal.pone.0255073.t001

Remote sensed covariates

We obtained remote sensed data on drought severity, malaria incidence, and population density. The description, and source to these data sets is provided in Table 2 .

https://doi.org/10.1371/journal.pone.0255073.t002

For example, the malaria incidence data was obtained from the Malaria Atlas Project (MAP). The project collects malaria data on malaria cases reported by surveillance systems, nationally representative cross-sectional surveys of parasite rate, and satellite imagery capturing global environmental conditions that influence malaria transmission [ 35 ].

Methodology

We assessed the relationships between socio-economic and remote sensed characteristics and anthropometric outcomes using the Bayesian Distributional Regression (BDR). BDR models all parameters of the response distribution based on structured additive predictors and allows to incorporate for example, non-linear effects of metric covariates, spatial effects, or varying effects. Applications of structured additive regression models to topics in Global Public Health are found in several publications [ 39 – 42 ]. This approach permits us to fully analyse the whole distribution [ 41 , 43 ] and our analysis was not restricted to assessing the conditional mean of the height-for-age z-score. Instead suspected heterogeneity across socio-economic and georeferenced factors and the anthropometric measure can be directly captured. In the context of growth failures this is of particular importance, as previous studies highlighted high levels of heterogeneity related to growth failures [ 33 ].

Bayesian distributional regression

Relying on Bayesian distributional regression requires to specify the distribution of the response variable. Assuming the response distribution to be Gaussian permits to model besides the conditional mean also the variance or standard deviation of the response variable. Graphical analysis using amongst others randomised quantile residuals [ 44 ] strengthens that a Gaussian model is plausible. See also Fig 2 , for more details. In the left- hand panel of Fig 2 the histogram of the height-for-age z-score together with the underlying density illustrates why the normal distribution seems to be an appropriate choice. This is further confirmed in the second and third panel, where the histogram of the quantile residuals including the underlying kernel density estimate, respectively, the QQ-plot of the randomised quantile residuals are shown.

The left-hand panel shows the histogram and kernel density estimates of the height-for-age z-score, the middle panel shows the histogram of the randomised quantile residuals together with the normal density estimates, and the right-hand panels depicts the QQ-plot of the randomised quantile residuals. Source : Demographic and Health Surveys (data); calculation by authors.

https://doi.org/10.1371/journal.pone.0255073.g002

Model selection

The fit of the models are compared by relying on the Deviance Information Criterion (DIC) [ 50 ] and Widely Applicable Information Criterion (WAIC) [ 51 ], and are summarised in Table 3 . As a rule of thumb can be seen that the model with the lowest value describes the data best. We specified six distinct models, aiming to identify the importance of, for instance, socio-economic or georeferenced factors. In more detail, the differences between these models are summarised in Table 4 .

https://doi.org/10.1371/journal.pone.0255073.t003

https://doi.org/10.1371/journal.pone.0255073.t004

In the following Section we will discuss the results of Model 5, omitting insignificant terms, as Model 5 has both the lowest DIC and WAIC. Result of the included covariates are however, similar throughout all specifications.

Descriptive analysis

Table 5 shows the baseline characteristics of selected covariates in the population between the two ZDHS survey of 2007 and 2013/14, and remote sensed data aggregates for these waves.

https://doi.org/10.1371/journal.pone.0255073.t005

Data was available for 25,852 children from the two waves, 31% observation were from the 2007 and 69% were from the 2013/14 ZDHS. Levels of stunting decreased between 2007 and 13/14 from a mean z-scores of -1.59 CI(-1.63; -1.55) to -1.47 CI(-1.49; -1.44). The breastfeeding duration declined from 16.22 to 15.69. There was a notable increase in the number of received vaccinations by children from 5.6 to 7.5 vaccinations. There was a slight increase in the number of years the mother spent in school from 7.2 to 7.8. Malaria incidence rates (plasmodium falciparum incidence) declined from 26% to 20%. Night-time light increased from 2.75, to 3.72 (observed values were log transformed), a possible indication of increase in urbanisation. Night-time light was highly correlated ( ρ = 0.73) to population density as such it was omitted in subsequent analysis.

High disparities in the height-for-age z-score have been observed at the district and provincial level within Zambia. There was a drift in the spatial pattern of malnutrition in the 2013/14 wave compared to the previous survey, indicating a general improvement. See also Fig 1 for a more detailed, descriptive, analysis of the spatial patterns of the height-for-age z-score within Zambia.

We observed that in the 2007 wave, stunting was lowest in the Western and Muchinga province. In the Southern province generally, low values were also observed, except for the Sinazongwe district. For Eastern province, Nyimba, Katete, Petauke and Lundazi districts had high levels. In the Luapula province, high levels of stunting were observed in the districts of Milenge, Mwense, Kawambwa, Nchelenge and Chiengi. Stunting was severe in some parts of the Copperbelt province which is predominantly a mining region and the Northern province. Central province had moderate levels, except for Serenje district.

Linear effects

With respect to the linear effects, Table 6 shows the effect of the gender and the area of residence on the posterior mean of the response variable. Considering the posterior mean of the height-for-age z-score of -1.70, boys were found to more stunted compared to girls. Stunting was also found to be higher in children from rural households compared to urban areas. Two patterns well documented in the literature for other countries and also Zambia [ 12 , 17 , 52 ].

https://doi.org/10.1371/journal.pone.0255073.t006

Socio-economic characteristics

Fig 3 shows the non-linear effect of the asset index, and the years of education of the mother on the mean μ and standard deviation σ of the response variable. Undernutrition has been associated to poverty [ 4 ], we observed that children living in poor household showed worse outcomes compare to children living in wealthier households, i.e the z-score is linearly increasing with increasing asset index. The effect of the asset index on the standard deviation does not notably vary across the range of the asset index, indicating a homogenous effect of wealth.

The Figure depicts the mean effects on the mean μ (left), and the standard deviation σ (right) together with 80 per cent and 95 per cent simultaneous credible intervals for the asset index (top), and the years of education of the mother (bottom). Source : Demographic and Health Surveys (data); calculation by authors.

https://doi.org/10.1371/journal.pone.0255073.g003

The bottom panel of Fig 3 shows that an increase in the level of education of the mother above eight years of education is associated with an appreciable increase in the height-for-age z-score. Notably also is that, there is not much difference in this effect for mother’s education less than 8 years. This entails that primary school education does not improve the nutrition outcome of the children as much. Above eight years of schooling, we see clearly that increase in the years has positive effect on the z-score. Moreover, the variation in the height-for-age z-score is higher with less years of schooling, whereas the variation gradually decreases with increasing levels of education of the mother.

Fig 4 shows the non-linear effect of the number of vaccinations the child received and mother’s BMI on the mean and the standard deviation z- score. The top graph shows that there was a positive effect on the mean z-score with increase in the number of vaccinations the child received. There is also greater variation in stunting levels among children who received less than 2 vaccinations.

The Figure depicts the mean effects on the mean μ (left), and the standard deviation σ (right) together with 80 per cent and 95 per cent simultaneous credible intervals for the vaccination coverage (top), and the BMI of the mother (bottom). Source : Demographic and Health Surveys (data); calculation by authors.

https://doi.org/10.1371/journal.pone.0255073.g004

Low values of the mothers BMI are negatively associated with the height-for-age z-score of the child, while for increasing values of the BMI also an increase in the posterior mean of the z/score can be observed. For values above 40 for the BMI of their mother the results are inconclusive indicated by the widening of the credible intervals. Low values of the BMI of the mother are associated with less variation compared to high values.

Fig 5 shows that increasing malaria incidence about 0.3 had a negative effect on the z-score, however we do not see any meaningful differences in the standard deviation over the spectrum the malaria incidences.

The Figure depicts the mean effects on the mean μ (left), and the standard deviation σ (right) together with 80 per cent and 95 per cent simultaneous credible intervals for the malaria incidence. Source : Demographic and Health Surveys (data); calculation by authors.

https://doi.org/10.1371/journal.pone.0255073.g005

Due to the high correlation of breastfeeding and age of the child, an interaction between these two variables can be presumed for which one has to account for. Fig 6 shows that children below 12 months of age who were breastfeed, were not malnourished. Accordingly, malnutrition mostly seems to be a process that comes to effect as children grow. Stunting was high for children above 36 months of age and who were breastfeed. On the right side, the figure shows that stunting was low in children whose mothers were around 30 years and with respect to birth order, which emphasizes that especially children of very young mothers are those most vulnerable.

The Figure depicts the mean effects on the mean μ for the interactions of the age of the child and breastfeeding duration, and the interaction the age of the mother and the birth order, respectively. Source : Demographic and Health Surveys (data); calculation by authors.

https://doi.org/10.1371/journal.pone.0255073.g006

Georeferenced characteristics

Fig 7 emphasizes the pronounced north and south pattern after adjusting for all the other variables, in particular after adjusting for wealth and rurality which was already described in the descriptive analysis. The highest variation in the height-for-age z-score was also observed in north (Luapula province) in both waves as can be seen on the right side of the figure.

The Figure depicts the mean spatial effect of the mean μ (left), and the standard deviation σ (right) for the year 2007 (left) and the year 2013/14 (right). Source : Demographic and Health Surveys (data) and Database of Global Administrative Areas (boundary information); calculation by authors. The shapefile used to create these maps is republished from [ 54 ] under CC BY license, with permission from Robert J. Hijmans, original copyright [2021].

https://doi.org/10.1371/journal.pone.0255073.g007

Discussion & conclusion

Using the two waves of the ZDHS, we modelled the height-for-age z-score by using socio-economic and remote sensed information. To analyse the whole distribution and not just focusing on the conditional mean, we used a Bayesian distributional regression approach accounting for heterogeneity as well.

Using Bayesian distributional regression, we assessed the relationship of socio-economic, and remote sensed covariates and stunting. Bayesian distributional regression, presents an advantage in terms of model flexibility allowing to incorporate, amongst others, non-linear effects and spatial effects. This however comes also with the drawback of data intensity and computational complexity.

Remote sensed techniques can be useful for future research on community health assessment as these techniques provide an advantage to take measurements quickly for remote and hard to reach areas. The data also enable to make meaningful analyses at sub-national levels which can improve targeting of interventions due to high levels of geographic specificity [ 26 , 33 ], however they do not give a full picture. Therefore, it is important to account for other covariates such as socio-economic characteristics at the individual or household level.

When relying on remote sensed information to asses anthropometric measures or biophysical developments, great caution should be taken with respect to data quality. Our study finds that remote sensed covariates alone explain little of the variation of the response, this emphasizes the need to control also for socio-economic characteristics. We find that the combination of remote sensed data and socio-economic characteristics explain more of the variation of the response, compared to solely focusing on one of the two sources of explanatory variables. In addition this also highlights the strong influence of socio-economic covariates or can be seen as an indicator of poor quality of the available remote sensed information.

Clear non-linear patterns emerged with respect to the years of education of the mother, and number of vaccinations. There was a clear non-linear tendency among children whose mothers had up to eight years of schooling having a low height-for-age z-score. For children of mothers with secondary or higher education the height-for-age z-score starts to improve strongly. This trend is consistent with what has been observed in others studies were odds of stunting were higher among children from mothers who had few years of education [ 12 , 52 ] and lowest among those who had advanced years in education [ 52 ]. Higher education level has been associated with increased income levels and improved knowledge among mothers who are usually the primary caregivers. As such educated mothers are more likely to take better care of their children by making informed nutritional decisions [ 24 , 52 , 53 ]. Increasing number of vaccinations showed improved z-score among children. Even though the effect was significant, the same size of the effect might not be relevant in practice.

Moreover, considering the full distribution like we did shows that the variation is highest for low levels of education and decreases with increasing years of education. This study did not consider the association of paternal education and the child z-score, however in another study, it was found that it was Maternal education that had a positive impact on children’s nutritional status [ 17 ].

We observe differences in levels of malnutrition in various regions in Zambia. One consistent pattern is that of discrepancy between the rural areas which are worse off compared to urban areas and confirms socio-economic inequalities between rural and urban areas. This may suggest social and economic inequalities between such areas. This has already been documented in other studies [ 12 , 14 , 24 ]. Furthermore, in terms of a spatial distribution, when you consider a smooth spatial effect, there is a clear regional variation in addition to the effect of rurality. Even after accounting for economic activities, the farming southern regions tend to be well off compared to the more industrialised northern areas. There is need to investigate further the underlying factors that contribute to the variation in the height-for-age z-score.

The present study shows that stunting still remain high in Zambia with remarkable regional inequalities and the decline is gradual which is unacceptable. There is need therefore to address the socio-economic indicators if this status is to improve.

Supporting information

S1 fig. sampling paths..

The Figure depicts the sampling paths of the parameters in η μ of the effect of the asset index, the two-dimensional effect of the birth order and the age of the mother at birth, the two-dimensional effect of the age of the child and breastfeeding duration, and the effect of the mothers years of education. Demographic and Health Surveys calculation by authors.

https://doi.org/10.1371/journal.pone.0255073.s001

S2 Fig. Sampling paths.

The Figure depicts the sampling paths of the parameters in η μ of the household size, the linear effects, and the malaria incidence. Demographic and Health Surveys calculation by authors.

https://doi.org/10.1371/journal.pone.0255073.s002

S3 Fig. Sampling paths.

The Figure depicts the sampling paths of the parameters in η μ of the maternal BMI, the population density, the aridity index, and the spatial effect. Demographic and Health Surveys calculation by authors.

https://doi.org/10.1371/journal.pone.0255073.s003

S4 Fig. Sampling paths.

The Figure depicts the sampling paths of the parameters in η μ of the effect of the number of vaccination. Demographic and Health Surveys calculation by authors.

https://doi.org/10.1371/journal.pone.0255073.s004

S5 Fig. Sampling paths.

The Figure depicts the sampling paths of the parameters in η σ of the effect of the asset index, and the effect of the mothers years of education. Demographic and Health Surveys calculation by authors.

https://doi.org/10.1371/journal.pone.0255073.s005

S6 Fig. Sampling paths.

The Figure depicts the sampling paths of the parameters in η σ of the household size, the linear effects, and the malaria incidence. Demographic and Health Surveys calculation by authors.

https://doi.org/10.1371/journal.pone.0255073.s006

S7 Fig. Sampling paths.

The Figure depicts the sampling paths of the parameters in η σ of the maternal BMI, the population density, the aridity index, and the spatial effect. Demographic and Health Surveys calculation by authors.

https://doi.org/10.1371/journal.pone.0255073.s007

S8 Fig. Sampling paths.

The Figure depicts the sampling paths of the parameters in η σ of the effect of the number of vaccination. Demographic and Health Surveys calculation by authors.

https://doi.org/10.1371/journal.pone.0255073.s008

Acknowledgments

The computational results presented have been achieved (in part) using the HPC infrastructure LEO of the University of Innsbruck.

• 1. World Health Organization. Global nutrition targets 2025: Stunting policy brief. World Health Organization; 2014.
• View Article
• Google Scholar
• 3. World Health Organization. The state of food security and nutrition in the world 2018: building climate resilience for food security and nutrition. Food & Agriculture Organization; 2018.
• PubMed/NCBI
• 13. UNICEF. Strategy for Improved Nutrition of Children and Women in Developing Countries. A UNICEF Policy Review. New York: UNICEF; 1990.
• 15. UNICEF. Improving child nutrition: the achievable imperative for global progress. New York; 2013.
• 19. Central Statistical Office (CSO)[Zambia] Ministry of Health (MOH)[Zambia] and ICF International. Zambia demographic and health survey 2013–14; 2014.
• 23. Klein N. An Introduction to Bayesian Structured Additive Distributional Regression; 2013.
• 24. Kandala NB, Lang S, Klasen S, Fahrmeir L. Semiparametric Analysis of the Socio-Demographic and Spatial Determinants of Undernutrition in Two African Countries; 2001.
• 25. Belitz C, Hübner J, Klasen S, Lang S. In: Determinants of the Socioeconomic and Spatial Pattern of Undernutrition by Sex in India: A Geoadditive Semi-parametric Regression Approach. Heidelberg: Physica-Verlag HD; 2010. p. 155–179.
• 37. Barichivich J, Osborn TJ, Harris I, van der Schrier G, Jones PD. Drought. In: Hartfield G, Blunden J, Arndt DS, editors. State of the Climate in 2018. vol. 100. Bulletin of the American Meteorological Society; 2018.
• 38. Center for International Earth Science Information Network—CIESIN—Columbia University. Gridded Population of the World, Version 4 (GPWv4): Population Density, Revision 10; 2017.
• 45. Belitz C, Brezger A, Klein N, Kneib T, Lang S, Umlauf N. BayesX: Software for Bayesian Inference in Structured Additive Regression Models. Version 3.0.2; 2015. Available from: http://www.BayesX.org/ .
• 49. Fahrmeir L, Kneib T, Lang S, Marx B. Regression: Models, Methods and Applications. 1st ed. Springer-Verlag Berlin Heidelberg; 2013.
• 54. Hijmans, Robert J, University of California, Berkeley, Museum of Vertebrate Zoology. Global Administrative Areas (GADM), version 3.6; 2018. Available from: http://gadm.org .

Contextual factors and spatial trends of childhood malnutrition in Zambia

Affiliations.

• 1 Department of Population Studies, University of Zambia, School of Humanities and Social Sciences, Lusaka, Zambia.
• 2 Department of Demography and Population Studies, Schools of Public Health and Social Sciences, University of the Witwatersrand, Johannesburg, South Africa.
• 3 Ministry of Mines and Minerals Development, Lusaka, Zambia.
• 4 University of Global Health Equity (UGHE), Bill and Joyce Cummings Institute of Global Health, Institute of Global Health Equity Research (IGHER), Kigali, Rwanda.
• 5 Institute of Global Health Equity Research (IGHER), University of Global Health Equity, Kigali, Rwanda.
• 6 Ignite Global Health Research Lab, Global Research Institute, William and Mary, Williamsburg, Virginia, United States of America.
• PMID: 36327254
• PMCID: PMC9632925
• DOI: 10.1371/journal.pone.0277015

Background: Understanding the national burden and epidemiological profile of childhood malnutrition is central to achieving both national and global health priorities. However, national estimates of malnutrition often conceal large geographical disparities. This study examined the prevalence of childhood malnutrition across provinces in Zambia, changes over time, and identified factors associated with the changes.

Methods: We analyzed data from the 2013/4 and 2018 Zambia demographic and health surveys (ZDHS) to examine the spatial heterogeneity and mesoscale correlates of the dual burden of malnutrition in children in Zambia. Maps illustrating the provincial variation of childhood malnutrition were constructed. Socio-demographic and clinical factors associated with childhood malnutrition in 2013 and 2018 were assessed independently using a multivariate logistic model.

Results: Between 2013/4 and 2018, the average prevalence of stunting decreased from 40.1% (95% CI: 39.2-40.9) to 34.6% (95% CI:33.6-35.5), wasting decreased from 6.0% (95% CI: 5.6-6.5) to 4.2% (95% CI: 3.8-4.7), underweight decreased from 14.8% (95% CI: 14.1-15.4) to 11.8% (95% CI: 11.2-12.5) and overweight decreased from 5.7% (95% CI: 5.3-6.2) to 5.2% (95% CI: 4.8-5.7). High variability in the prevalence of childhood malnutrition across the provinces were observed. Specifically, stunting and underweight in Northern and Luapula provinces were observed in 2013/14, whereas Lusaka province had a higher degree of variability over the two survey periods.

Conclusion: The study points to key sub-populations at greater risk and provinces where malnutrition was prevalent in Zambia. Overall, these results have important implications for nutrition policy and program efforts to reduce the double burden of malnutrition in Zambia.

• Growth Disorders / epidemiology
• Malnutrition* / epidemiology
• Risk Factors
• Thinness* / epidemiology
• Zambia / epidemiology

Grants and funding

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Int J Environ Res Public Health

Community-Based Management of Child Malnutrition in Zambia: HIV/AIDS Infection and Other Risk Factors on Child Survival

Stefania moramarco.

1 Department of Biomedicine and Prevention University of Rome Tor Vergata, via Montpellier, Rome 00133, Italy; [email protected] (S.M.); [email protected] (L.P.); [email protected] (E.B.)

2 Rainbow Project Association Pope John 23rd, 5656 Chinika Road, Ndola 10101, Zambia; [email protected] (G.A.); [email protected] (C.C.)

Giulia Amerio

Clarice ciarlantini, jean kasengele chipoma.

3 Ndola District Health Office, 1307 Naidu Close, Ndola 10101, Zambia; moc.liamg@elegnesaknaej (J.K.C.); ku.oc.oohay@ewgnupmisugnukakm (M.K.S.)

Matilda Kakungu Simpungwe

Karin nielsen-saines.

4 Department of Pediatrics, David Geffen UCLA School of Medicine, Los Angeles, CA 90095, USA

Leonardo Palombi

Ersilia buonomo.

(1) Background: Supplementary feeding programs (SFPs) are effective in the community-based treatment of moderate acute malnutrition (MAM) and prevention of severe acute malnutrition (SAM); (2) Methods: A retrospective study was conducted on a sample of 1266 Zambian malnourished children assisted from 2012 to 2014 in the Rainbow Project SFPs. Nutritional status was evaluated according to WHO/Unicef methodology. We performed univariate and multivariate Cox proportional risk regression to identify the main predictors of mortality. In addition, a time-to event analysis was performed to identify predictors of failure and time to cure events; (3) Results: The analysis included 858 malnourished children (19 months ± 9.4; 49.9% males). Program outcomes met international standards with a better performance for MAM compared to SAM. Cox regression identified SAM (3.8; 2.1–6.8), HIV infection (3.1; 1.7–5.5), and WAZ <−3 (3.1; 1.6–5.7) as predictors of death. Time to event showed 80% of children recovered by SAM/MAM at 24 weeks. (4) Conclusions: Preventing deterioration of malnutrition, coupled to early detection of HIV/AIDS with adequate antiretroviral treatment, and extending the duration of feeding supplementation, could be crucial elements for ensuring full recovery and improve child survival in malnourished Zambian children.

1. Introduction

Zambia is a sub-Saharan country facing a high burden of child acute malnutrition, with malnutrition remaining one of the most serious problems among children under five years of age [ 1 ]. According to the Zambian Preliminary Report of Demographic and Health Survey 2013–2014, 40% of children are affected by stunting, 15% are underweight, and 6% of children suffer from wasting, with a high under-five mortality rate (75 deaths per 1000 live births in a year) [ 2 ]. Malnourished children who do not quickly break away from the vicious cycle of infectious disease and growth failure are vulnerable to irreversible cognitive damage [ 3 ]. Acute malnutrition (wasting), especially in severe form, if untreated, is an attributable cause of the 12.6% of the 6.9 million deaths worldwide among children under five years of age [ 4 ]. In Zambia, malnutrition has been estimated to underlie up to 52% of all under-five deaths [ 5 ]. The picture of malnutrition is exacerbated by the HIV/AIDS pandemic: when the condition of being HIV-positive coexists with malnutrition, the risk of growth failure and morbidity increases, and children delay recoveries and suffer relapses of malnutrition events [ 6 , 7 ], with a longer need of nutritional rehabilitation therapy compared to their HIV-negative counterparts [ 8 ]. A most recent research demonstrates that acute and chronic malnutrition decrease the odds of attaining adequate motor milestones even in HIV-exposed, but not infected, children [ 9 ]. The Zambian Government has been committed to address child malnutrition, by implementing important health sector reforms aimed at strengthening health service delivery in order to improve the health of Zambian children [ 10 ]. The process started in 2005, with the community-based therapeutic care (CTC) for the management of malnutrition first introduced in Lusaka district [ 11 ]. The community-based management of malnutrition has been endorsed at the international level as an innovative cost effective approach for managing child malnutrition, integrated into the routine health system of many African countries, preventing hundreds of thousands of child deaths when applied at scale [ 12 ]. This model of intervention is combining facility- and community-based approaches, which reserve inpatient care (IC) based on the World Health Organization (WHO) recommended management protocol for malnourished children with medical complications [ 13 ], and outpatient therapeutic care (OTP) complemented by supplementary feeding programs (SFP) for uncomplicated cases [ 14 ]. Targeted supplementary feeding programs (SFPs) are particularly effective in the management of acute malnutrition, treating moderate acute malnutrition (MAM) and preventing the deterioration into severe acute malnutrition (SAM); however, the management of MAM is still debated and consensus regarding management has still not been reached, particularly in non-emergency situations and relatively food-secure settings in low and middle-income countries [ 15 ]. SFPs in Zambia are not widely available, and most areas are not covered by nutrition-specific interventions targeting MAM. Rainbow Project SFPs are the only well implemented sites with this specific approach in the Copperbelt. There is a need for more evidence-based data reporting on the effectiveness of SFPs, not only to address areas of weakness and plan new interventions, but also to support the potential need for greater implementation of SFPs in Zambia, taking into consideration different scenarios.

2. Materials and Methods

2.1. setting.

The current study presents data on malnourished Zambian children assisted in the Rainbow Project SFPs. The Rainbow Project, under the Pope John 23rd Association, is a large-scale model of care for orphans and vulnerable children operating since 1998 in the Ndola and Kitwe Districts. The model is made by several components, including the community-based program against malnutrition. In the context of a traditional CMAM, Rainbow Project is running SFPs with a particular focus on the community mobilization and capacity building activities. In the Ndola District, 11 SFPs are present in different areas around the city. All of the centers are run by leaders of ten small non-governmental organizations (NGOs) and community-based organizations (CBOs), and are coordinated by professionals of Rainbow office, in close network with Ndola District Health Management Teams (DHMTs) and other local authorities. All the SFPs activities are performed by community volunteers/operators appropriately trained in Infant and Young Child Feeding (IYCF) practices promoted by the Zambian Government and who are constantly updated on nutrition topics. Personnel are also trained in confidentiality issues (e.g., counseling in the context of prevention of mother-to-child transmission PMTCT) [ 16 ].

2.2. Malnutrition Classification, Management and Integration of CMAM Components

Malnourished children enrolled in Rainbow SFPs were recruited through community outreach or referred from local health facilities. Children were admitted to SFPs by using a two-priority criteria system of enrollment: first priority was given to acute malnutrition (SAM or MAM) and second priority to underweight status. If a child qualified at the same time for different criteria, the enrollment was made considering the most severe condition of malnutrition. According to WHO/UNICEF criteria [ 17 ] and the Integrating Management of Acute Malnutrition (IMAM) guidelines of the Zambian Ministry of Health [ 18 ] for children aged 6 to 59 months, SAM was defined as a weight-for-height (WHZ) or a weight-for-length z-score (WLZ) ≤ −3, mid-upper arm circumference (MUAC) ≤ 11.5 cm, and/or presence of bilateral pitting edema (kwashiorkor); MAM was defined as WHZ/WLZ < −2 and >−3, MUAC ≤ 12.5 cm and >11.5 cm. As recently recommended by the updated WHO guidelines on the management of severe acute malnutrition in infants and children, in order to achieve early community identification of malnourished children, Rainbow’s well-trained volunteers measured the MUAC and examined children for pitting edema, allowing the assessment of WHZ/WLZ to be done by health workers within primary health care facilities and hospitals [ 19 ]. According to the WHO standard, underweight was defined as a weight-for-age z-score (WAZ) < −2 [ 20 ]. Children identified as having SAM were referred to the Arthur Davidson Children’s Hospital for inpatient treatment if medical complications were present, or to the OTP in the absence of medical complications. All children with MAM were eligible for SFPs: children with MAM without health complications were directly enrolled, while children with MAM with health complications were first to be referred to the nearest health facility for immediate medical care. For ethical and humanitarian reasons, Rainbow SFPs admitted children with SAM or health complications despite additional referral for admission to the nearest health facility for proper screening and advice from the health staff. SFP admission was made possible when CMAM was not fully implemented or access to the hospital was restricted (i.e., OTPs were not covering the whole Ndola area, the availability of ready-to-use therapeutic food/RUTF was erratic, inpatient care was not provided or not applicable), in order to facilitate access to food for more severe cases.

2.3. SFP Activities and Data Collection

On a weekly basis, the SFP protocol included anthropometric follow-up, data recording, meals in loco (porridge meal), nutritional supplementation, home visits, as well as health talks and cooking demonstrations for mothers/care-givers. Anthropometric follow up consisted in measuring weight, MUAC, and checking for bilateral pitting edema. Children were assessed without clothes or footwear; weight (in kilograms) was measured using a mechanical baby scale graduated by 0.1 kg increments (salter 235). Mid-upper arm circumference (in centimeters) was measured using a simple colored plastic strip (standardized UNICEF tape). Bilateral pitting edema was checked by applying gentle thumb pressure on the dorsum of the feet and assessing for residual depression; edema was detected as different grades.

Data was collected on general and socio-demographic characteristic, as well as health and nutritional conditions. All information was entered in a register edited ad hoc by the Rainbow Project, in order to monitor health and nutritional status of the beneficiaries. General demographic pediatric information included date of birth, age, gender, and siblings. Socio-demographic data recorded included family history (parents’ marital status, relationship and age of the guardian), and housing information (area of stay, address, household conditions). Health information included disability, medical complications and/or illness, enrollment in OTP, and HIV status. The latter (HIV status), was ascertained from the PMTCT section of the child under-five card released from the primary health care facilities of the Zambia Health Ministry. Nutritional parameters included anthropometric measurement (weight, MUAC, edema). Data was collected after verbal consent of caregivers and in full respect of confidentiality.

Educational activities for mothers/guardians included provision of health talks and cooking demonstrations. A meal in loco was offered to all of the children attending the program. It mainly consisted in porridge, prepared with a high-energy protein supplement (HEPS). The high-energy protein supplement is a specific corn-soya blended food (CSB), fortified with micronutrients (vitamins and minerals), mostly recommended for the management of malnutrition in SFP. Children received on a weekly basis a food ration to take home as nutritional supplementation. Considering that SFPs were conducted in food insecure areas where availability of food was generally limited, local food for the whole household was distributed in addition to HEPS, so that the specific supplement for the child was not shared within the family. Children with SAM without medical complications, if enrolled in the OTP, received RUTF from health facilities, when available.

Home visits were performed by community volunteers to ensure compliance with nutritional and health guidelines. All children stayed in the program until SFP discharge criteria were met: for two consecutive weeks edema should be absent and MUAC > 12.5 cm, or 15% weight gain had to be considered if underweight was the admission criteria.

2.4. Study Population, Statistical Analysis, and Ethical Considerations

This was a community-based retrospective observational study. Pediatric data was routinely collected and entered in the database with removal of personal identifiers. The current study presents data on malnourished Zambian children aged 6 to 59 months who were assisted from 2012 to 2014 in the Rainbow Project SFPs around the Ndola area. A database with all pediatric records coming from different SFPs was generated. Data were extracted and analyzed using SPSS software system 20.0 (IBM, Somers, NY, USA). Weight-for-age z-scores (WAZ) were calculated using the WHO Anthro Software (Version 3.2.2, January 2011, WHO, Geneva, Switzerland), based on the 2006 World Health Organization Child Growth Standards [ 20 ]. Descriptive data and variables measured were presented as means with standard deviations (SD). For recovered children differences between means of anthropometric parameters were tested with the student t -test. The odds ratios, 95% confidence intervals, between age and length of stay, age and SAM, HIV, and MAM were calculated. Univariate and multivariate Cox regression (forward stepwise model) was performed to identify the main predictors of mortality and cure.

The study protocol was approved by the Tropical Diseases Research Centre (TDRC) Ethics Committee of Ndola, Zambia (IRB registration number 00002911).

2.5. Program Outcomes and Performance Indicators

In order to evaluate program performance, exit categories for targeted SFPs from Sphere Project and UNHCR guidelines were considered. The Sphere Project defines the standards by which the international community responds to the plight of people affected by disasters, principally through a set of guidelines that are set out in the Humanitarian Charter and Minimum Standards in Humanitarian Response. Sphere standards are the typical criteria used for assessing the effectiveness of SFP [ 21 ]. UNHCR guidelines are intended as a practical guide to design, implement, monitor, and evaluate selective feeding programs in emergency situations [ 22 ].

Standard outcomes were defined as recovery rate, death rate, and default rate. Recovered/cured was defined as an individual who met the discharge criteria. Defaulter was defined as a child lost to follow up for three consecutive weeks. A child was classified as “defaulter” when he/she dropped out of the study due to refusal or it was not possible to locate the child and make a home assessment. Death was registered when occurring during the time the patient was enrolled in the program. Early mortality (within 15 days from enrollment) was excluded because it might not be directly attributable to the performance of the SFPs. Individuals who did not complete their treatment because they moved to another area were considered transferred. This outcome was not included in the performance evaluation because of the current absence of published targets. Length of stay and weight gain were considered additional indicators for targeted SFPs [ 23 ]. Mean length of stay expressed the average time of stay for recovered children; mean weight gain expressed the average number of grams gained per kg per day among children who were cured. For humanitarian and ethical reasons treatment was provided until children reached the recovery goals (treat-to-goal), so none were categorized as non-cured/non-responder (defined as cases that did not reach discharge criteria after a pre-defined length of time). We reviewed the literature in order to compare Rainbow Project’s performance with published outcomes of other similar programs.

Information about SFPs within the Ndola area were considered for this study. Specifically, we focused on 10 SFPs, eight operating in urban areas (Twapia, Nkwazi, Kabushi, Kaloko, Kawama, Chifubu, and Mackenzie), and two located in rural areas (Baluba and Chikumbi); only one center was not entered in the database. Data for 1226 children, all coming from low socio-economic households and assisted from 2012 to 2014, were extracted from the database. Children still on SFPs treatment at the moment of the study were not included. Formally transferred and early mortality episodes were excluded. Twelve cases were left out because of missing data. A total of 858 children (49.9% male) provided data for the analysis ( Figure 1 ).

Study flowchart.

Table 1 reports the demographic, social, health, and nutritional characteristics of children at baseline. The median age was 19 months ± 9.4 SD, with a range from 5 to 60 months. More than half of the sample was less than 18 months of age (53%). Children younger than 18 months were more likely to be severely malnourished (OR 1.5 CI 1.1–2.1). Regarding the admission criteria, 37.6% of children were affected by MAM, 28.1% were affected by SAM, and 34.3% were admitted due to being underweight. Bilateral pitting edema was identified in 9% of the children. The mean weight of the sample was 7.8 kg ± 1.6 SD; the mean MUAC was 12.4 cm ± 1 SD; the mean WAZ was −2.8 ± 1.1 SD. Sixty-eight children (7.9%) suffered relapses of malnutrition (defined as a new episode of malnutrition after previous discharge). At time of admission, health problems affected 34.9% of beneficiaries; specifically 8.5% had fever, 8.3% had diarrhea, 7.7% had lack of appetite, 3.3% had cough, 0.8% had malaria, and 6.3% had other unspecified conditions. As measures of health problems, we asked the mother/caregiver if the child was experiencing any illness at the moment of the enrollment in the SFP. No physical examinations were performed.

Demographic, social, health, and nutritional characteristics of children at baseline.

* data available since May 2014.

Figure 2 shows HIV status at admission and discharge. At enrollment, 51 children were reported as HIV infected (5.9%), 426 HIV uninfected (49.7%), while for 381 children the HIV status was unknown (44.4%). Supporting voluntary counseling and testing (VCT) was an SFP activity: all guardians of children with unknown HIV status were encouraged to go to the nearest health facility for an HIV rapid test for both mother and child; HIV-positive children not receiving ARV treatment were referred to the ART clinic for assessment of ART eligibility. At the time of discharge, 63 children (7.3%), that is 12 more children, were found to be HIV infected with 33 (52.3%) initiating ARV treatment, and six (18.2%) also initiating TB treatment; 580 children (67.6%) were confirmed to be HIV-negative, while still 215 children (25.1%) had an unknown HIV status. For only 2.2% HIV exposure was ascertained from the health cards. For the others it was not even possible to know the exposure to HIV.

Diagnosis of HIV status from enrollment to discharge.

3.1. Performance of the Rainbow Project Based on International Standards

In Table 2 outcomes of the Rainbow Project, reported as for total, MAM, and SAM, were compared with international Sphere standards and UNHCR guidelines for the community-based management of acute malnutrition.

Rainbow Project SFPs outcomes reported as for total, MAM and SAM, and compared with International Standards.

* Sphere Project [ 20 ]; § UNHCR [ 21 ]; TFPs = Therapeutic feeding programs.

Outcomes for the three main core performance indicators (recovery, death, and defaulter rate) were compared with Sphere standards. The overall recovery rate was 82.6%; respectively, 86.1% for MAM and 73.5% for SAM. The global default rate was 11.8%, with 11.1% MAM and for 14.1% for SAM, all below the standard. The overall case-fatality rate was 5.6%, above the Sphere standard. When splitting the two groups, the death rate for MAM was below the target (2.8%), while the case-fatality rate for SAM was 12.4%. The number of deaths, which is above the alarming target was mainly due to the very critical health and poor nutrition conditions of these children associated with the low access to adequate health care. Two other targeted SFP indicators, length of stay and weight gain, were considered. The mean length of stay was 19.3 weeks; compared to International Standards, an increased number of weeks was needed for recovery of children with MAM (+7 weeks), and an even longer period for children with SAM (+10 weeks). Children younger than 18 months of age were more likely to stay longer (OR 1.4 95% CI: 1.1–1.8). The average weight gain was generally 1.7 g/kg/day. The mean weight gain for MAM was 1.7 g/kg/day, below the international guidelines (UNHCR). Published data reports a desirable weight gain of ≥5 g/kg/day [ 24 ], while it tended to be less in some studies (3 g/kg/day) [ 25 ] and reviews of the literature (between 1 and 2 g/kg/day) [ 26 ]. The mean weight gain of severely-malnourished children (2 g/kg/day) was far below that of UNHCR guidelines and recent literature (4.4 g/kg/day) [ 27 ].

3.2. Anthropometric Analysis of Recovered Children

In order to assess the nutritional status of recovered children, differences between anthropometric parameters’ means at admission and discharge (weight, WAZ, MUAC) were compared using Student’s t -test. Table 3 shows significant improvements ( p < 0.0001) on the anthropometric parameters: the weight gained from 7.8 kg ± 1.6 to 9.2 kg ± 1.6 (+1.4 kg ± 0.8); the WAZ rose from −2.8 ± 1.1 to −1.9 ± 0.9 (+0.8 ± 0.8); and the MUAC increased from 12.4 cm ± 1 to 13.7 cm ± 0.8 (+1.3 cm ± 0.9).

Differences between means of the anthropometric parameters at admission and discharge in recovered children ( n = 709), and Student’s t -test.

3.3. Predictors of Mortality and Cured

We performed a univariate and multivariate (forward stepwise model) Cox proportional risk regression to identify the main predictor of mortality ( Table 4 ). Variables of severity of malnutrition, HIV status, weight gain, weight for age z-score at admission, age in months and frequency of health problems, and health problems at admission were included in the analysis. Both analyses showed a significant association with SAM, HIV infection, and WAZ < −3 at admission.

Predictors of mortality. Cox proportional risk analysis.

Figure 3 shows the result of Cox survival analysis (outcome: death) by HIV status.

Cox survival analysis (outcome: death) by HIV status.

Finally, we performed a Cox analysis using as the dependent variable the “cured” status, calculating the time to event as seen in Table 5 .

Predictors of failed cured status. Cox proportional risk analysis.

Figure 4 shows time to event (cured children). At 24 weeks 80% of children result recovered by severe/moderate malnutrition.

Time to cure events as 1- survival.

4. Discussion

The main objective of this study was to evaluate the effectiveness of the model in the community management of child malnutrition in the Ndola district. Rainbow Project SFPs in the Ndola area assisted children from 6 to 59 months old with MAM and, for humanitarian and ethical reasons, children with SAM and/or health complications when CMAM was not fully implemented. Rainbow SFP activities included both nutrition-specific or direct interventions (growth monitoring and supplementary food) and nutrition-sensitive or indirect interventions (HIV counseling and testing, nutrition, and health skills for mother and child health promotion). General outcomes showed good program performance. Outcomes for MAM and SAM were analyzed separately in order to better compare the program performance with International Standards. Results of this study demonstrate that Rainbow SFPs is very effective in the management of MAM. Outcomes for MAM met all the Sphere standards for targeted SFPs, with high recovery rates (86.1%), and both low defaulter (11.1%) and mortality rates (2.8%). Comparing with other studies, our findings showed outcomes similar to other community-based programs for moderate malnutrition, or in some cases with a better cure rate but higher mortality rates [ 25 , 27 , 28 ]. Outcomes for children with SAM exceeded the international standards. Our findings reflected the most critical condition of severely malnourished children, for whom therapeutic care must be provided, especially when health complications coexisted. These findings were in accordance with a study on community-therapeutic care in Lusaka reporting a mortality for children having SAM treated with RUTF of more than 9%, while in the absence of treatment the mortality rate was expected to be above the 20% [ 11 ].

In order to better understand the reasons linked to the fatality rate, we identified a higher risk of mortality, respectively, for severe acute malnutrition, HIV infection, very low weight-for-age z-score at admission, and health problems at admission. These variables were all independent and strong predictors of mortality. With respect to HIV status, our analysis reported that HIV-infected children were almost three times more at risk of death than HIV uninfected. These results, according to Munthali and colleagues, demonstrate that Zambian malnourished children who are HIV-infected are 80% more likely to die than those who are HIV-uninfected [ 29 ], highlighting the role of the vicious cycle between HIV/AIDS infection and childhood malnutrition and its negative synergic impact on mortality. In setting where HIV infection is common (in Zambia 100,000 children aged 0 to 14 years are living with HIV [ 30 ]), the management of HIV/AIDS malnourished children remains more critical, since nutritional rehabilitation must be essentially integrated with early diagnosis efforts and the adequate ARV therapy [ 31 ]. At the moment of data collection the country had adopted WHO Option A as the national ART program for HIV/AIDS treatment and prevention. HIV-positive children below 24 months were recommended to initiate ART irrespective of CD4 counts or WHO HIV clinical stage; children between 24 and 59 months were recommended to initiate ART with CD4 counts of ≤750 cells/mm 3 or % CD4+ <25%, irrespective of WHO HIV clinical stage, or with WHO HIV clinical stages 3 and 4, irrespective of CD4 counts. Zambian guidelines for option B+ were launched in 2013 and progressively implemented. These new guidelines recommend lifelong triple-combination ART for all confirmed HIV-infected children regardless of CD4 count and/or WHO clinical stage.

Finally, the Cox analysis showed that recovering from severe-moderate malnutrition was reached at 80% of the population only at the end of month six. International guidelines recommend that the length of treatment of 12 weeks but our study shows that further benefits could be added by a prolonged treatment until week 24.

Despite the incisive work Rainbow SFPs performed in HIV counseling and testing, our analysis highlight high percentage of children who remained with an unknown HIV status at time of ending the program, underscoring that stigma HIV and fear of discrimination still bear in Zambia. We can presumably suppose that some of the children with HIV status unknown at the time of discharge from SFP could be HIV infected and, therefore, at more risk of morbidity and mortality. National guidelines must, therefore, be efficiently implemented and PMTCT programs must be improved, both at the community and at the health system level. Enhancing the integration among the different stakeholders dealing at local level with HIV and malnutrition could facilitate early detection of HIV infected children at most need of health care and support. Further research will also need to evaluate the positive health impact of both children and their family since shifting to the WHO 2013 guidelines in Zambia [ 32 ].

The low average weight gain suggests that the ration of high-quality food supplementation could be improved. A new food distribution consisting of redoubling the amount of HEPS could be suggested, in order to provide a daily ration of 150 g of HEPS per child. The ration of general local food would be still provided in order to support the household’s diet, considering the poor setting and food sharing within the nuclear family. Assuming that, to meet their daily dietary needs, children rely only on food supplements distributed by such programs, we estimate children could benefit from more than 1000 kcal/day, from either HEPS and local food, with a mean average of more than 25 g/day of protein. This new food schedule could meet the proposed recommended nutrient requirements for moderate malnourished [ 24 ], enhancing the effectiveness of the program. Although cost assessment was beyond the scope of this study, we have estimated the new food schedule to be sustainable, not affecting previous monthly costs for SFPs. Presumably, with a better weight gain due to the new food schedule, the mean length of stay for recovered children could decrease, improving the cost-effectiveness of the program in the long-term.

5. Conclusions

The Rainbow Project SFPs are effective and sustainable in the community management of child malnutrition in the Ndola district. Results from the Rainbow Project SFPs suggest that preventing deterioration in severe acute malnutrition, coupled to early detection of HIV/AIDS with adequate antiretroviral treatment, and extending the duration of feeding supplementation, are associated with an effective weight gain. These are crucial elements for ensuring full recovery and lowering mortality rates in malnourished Zambian children. In order to improve the Rainbow Project SFPs performance, a higher-quality food distribution was recently implemented. Meanwhile, enhanced education of community volunteers/operators on child malnutrition and HIV/AIDS knowledge has been reinforced.

Acknowledgments

We acknowledge with gratitude all the children and their guardians assisted in SFPs during these years of work. We thank the Association Pope John the 23rd and all the operators and volunteers of NGOs/CBOs involved in the nutritional program for their tremendous job in the community. We thank Elisa Magarotto who helped in data collection in the field and for merging data files used in these analyses. We are particularly grateful to Elisabetta Garruti and Gloria Gozza for project management, guidance and support. We would also like to thank all the Ndola District Health Office for the constant support and advice.

Abbreviations

The following abbreviations are used in this manuscript:

Author Contributions

Stefania Moramarco performed the statistical analysis and interpretation of the data, drafted and wrote the manuscript, made a substantial contribution to the local implementation. Giulia Amerio supervised the study at local level, contributed to the interpretation of the data, provided critical comment and revision of the manuscript, led the field implementation. Clarice Ciarlantini collected data, contributed to the revision of the manuscript and to the field implementation. Jean Kasengele Chipoma and Matilda Kakungu Simpungwe reviewed the paper and provided critical comment, made a substantial contribution to the local organization. Karin Nielsen-Saines reviewed the manuscript and contributed critical comments. Leonardo Palombi performed the statistical analysis, contributed to the data interpretation and provided critical comment and revision of the manuscript. Ersilia Buonomo supervised the study, performed the statistical analysis and the interpretation of the data, provided critical comment and revision of the manuscript. All authors red and approved the final manuscript.

Conflicts of Interest

The authors declare no conflict of interest.

Academia.edu no longer supports Internet Explorer.

To browse Academia.edu and the wider internet faster and more securely, please take a few seconds to  upgrade your browser .

Enter the email address you signed up with and we'll email you a reset link.

• We're Hiring!
• Help Center

The Status of Hunger and Malnutrition in Zambia: A Review of Methods and Indicators

Embassy of Sweden and the United States Agency for International Development (USAID) in Lusaka.

Related Papers

sandy findlay

Scott Drimie

Due to the predominance of direct, specific interventions in nutrition for development, the health sector tends to own nutrition, with interventions customarily implemented through health programs. The premise that the agriculture sector should also be a vehicle for improved nutrition is intuitive, but this sector often delivers neither good nutrition nor food security to the most vulnerable in the population. The complex and multisectoral nature of malnutrition may explain why it has not been effectively addressed. Though we know many of the solutions, such as intersectoral action which is critical to addressing this complexity, to date there is no consensus on how intersectoral solutions are best implemented or institutionalized. This review brings together experiences from across Sub-Saharan Africa in order to draw out recommendations for improved intersectoral implementation going forward, and assesses how these findings apply specifically to the Zambian context.

Tanzania journal of health research

Sophia Tembo

Background: Malnutrition is a severe and persisting cause of morbidity and mortality among under-five children in Zambia. We assessed malnutrition and its determinants among under-five children in Lusaka urban, Zambia. Methods: This cross-sectional study was conducted at Chawama, Kabwata, Chilenje and Mtendere urban clinics in Lusaka. A simple random sample of 384 under-five children were selected between February and March 2015. A structured interview schedule was used to collect data on demographic characteristics, determinants of malnutrition and anthropometric measurements. Multivariate binary logistic regression analysis was carried out using IBM® SPSS® Statistics for Windows Version 20.0 to identify determinants of malnutrition while adjusting for confounders. Results: Of the 384 under-five children, most (51.8%) of them were male and the mean (SD) age was 13.83±10.17 months. The zH/A; zW/H; and zW/A was 68.5% and 31.5%; 95.6% and 4.4%; and 86.2% and 13.8% for normal and stu...

Selestine Nzala

Zambia Social …

Dick Jonsson

Health Policy and Planning

Peter Delobelle

The evolution of nutrition patterns in Zambia has resulted in the coexistence of undernutrition and overnutrition in the same population, the double burden of malnutrition. While Zambia has strong policies addressing undernutrition and stunting, these do not adequately address food environment drivers of the double burden of malnutrition and the adolescent age group and hence the need for nutrition policy reforms. We conducted a theory-based qualitative prospective policy analysis involving in-depth interviews with nutrition policy stakeholders and policy document review to examine the feasibility of introducing nutrition policy options that address the double burden of malnutrition among adolescents to identify barriers and facilitators to such policy reforms. Using the multiple streams theory, we categorized the barriers and facilitators to prospective policy reforms into those related to the problem, policy solutions and politics stream. The use of a life-course approach in nutri...

Ernest Kakoma

Abstract Background: Poor maternal nutrition increases the risk of both maternal and child morbidity and mortality. In Zambia, the prevalence of undernutrition among women of reproductive age (WRA) is 10.3%. The determinants and interventions have not been studied. Objective: To explore determinants of maternal undernutrition, practices to address maternal undernutrition and recommend interventions in three provinces of Zambia. Methodology: Literature review was conducted using the 2013 Lancet framework for maternal and child undernutrition to guide the analysis of the determinants of maternal undernutrition. Findings: The determinants particularly affect the two rural provinces of Western and North western compared to more urban Lusaka province. They include high prevalence of poverty, early female adolescent pregnancies which affect education progression. Further, inadequate dietary diversity with high dependence on cereals and starchy roots coupled with low intake of micronutrients during pregnancy, high burden of infectious diseases like malaria in North western Province due to low insecticide treated bed nets (ITNs) distribution and high HIV prevalence in Lusaka and Western provinces. Low access to health services especially in rural provinces affects utilisation of services such as family planning, antenatal care and nutrition counselling, deworming and micronutrient supplementation receipt and adherence. Further, low access to sources of improved drinking water and sanitation in rural provinces results in increased risk of diseases, and micronutrient intake remains low due to limited dietary diversity and inadequate policies on nutrition and food fortification as only vitamin A is fortified in sugar and iodine in salt. Some interventions to address maternal undernutrition include iron-folic acid supplementation, deworming and provision of intermittent presumptive treatment of malaria including family planning for WRA. Some of the best practices to addressing maternal undernutrition include community and schools involvement. Conclusion: Maternal undernutrition has myriad of determinants resulting in inadequate dietary intake and increased burden of infections. Recommendations: Multisectoral approach supported by policies on nutrition and food fortification, conditional cash transfer, micronutrient supplementation for adolescent girls and pregnant women, tailored for supplementation for food insecure pregnant women, access to family planning including for adolescents, disease prevention and treatment, improved sources of drinking water and sanitation and research in nutrition and adolescents health. Key words: Maternal, undernutrition, determinants, interventions, Zambia Word count: 12,170

Kiersten Johnson

Mary Corbett

The author would like to acknowledge the hard work of the SHA team in Zambia, in particular SHA staff in Northern Province. Special recognition is given to Edward Meleki (Programme Coordinator) for all his support over the years. The author would also like to recognise and thank all other stakeholders involved in this programme, including the Ministry of Health and Ministry of Agriculture and Livestock (MAL). A special thanks to all the communities in Luwingu and Mbala districts. Finally, the author gives her sincere appreciation and thanks to Irish Aid for funding this innovative programme, giving the flexibility to SHA to adapt the programme to needs as they arose. Nutritionsensitive agriculture in Zambia: A continued work in progress Location: Zambia

Chipwaila Chunga

Background: Malnutrition continues to take the lives of millions of children every year. Children under the age of five years are especially susceptible to high rates of malnutrition and more than half of the deaths in under five children are caused by undernutrition in low- and middle-income countries. Zambia continues to have one of the highest rates of malnutrition in Southern Africa, specifically Severe Acute Malnutrition (SAM). The country’s prevalence of under-five malnutrition is above the WHO Public Health threshold which may linked to the observed under-mortality. In this study, we assessed the factors associated with under-5 mortality in children with SAM. Methods: This was a cross-sectional study assessing the factors associated with under-5 mortality in children hospitalized with severe acute malnutrition. The study used data from a cross-sectional study which was assessing factors associated with SAM in under-5 children in five underperforming provinces of Zambia. In th...

RELATED PAPERS

eduardo lopez

Afghanistan

Sara Peterson

Damiano Acciarino , Alessio Panichi , Valentina Prosperi

ismail gultepe

Paolo Divizia

yuniar soeli

Anugrah Ananda

Christian Jutten

International Journal of Sensors, Wireless Communications and Control

Mohuya Chakraborty

LINK 2021 Conference Proceedings

Rosangella Leote

Saeema Ahmed-Kristensen

做uwa学位证书 西澳大学毕业证学位证书留服认证原版一模一样

Lecture Notes in Computer Science

Carlo Emmanoel Tolla de Oliveira

Kost Elisevich

Journal of Applied Physics

Fernando Sabino

Las elecciones presidenciales en Chile ¿Una configuración ecuménica conservador

Mónica Ulloa Gómez

IEEE Signal Processing Letters

Arquivos Brasileiros de Endocrinologia & Metabologia

Josefina Bressan

Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering

Eugen Borcoci

Primary Health Care: Open Access

Nazan Savaş

Bet la destination ultime pour les jeux en ligne!

Brigitte Laliberté

Blucher Chemical Engineering Proceedings

Leandro Lissner

•   We're Hiring!
•   Help Center
• Find new research papers in:
• Health Sciences
• Earth Sciences
• Cognitive Science
• Mathematics
• Computer Science
• Academia ©2024

• Research article
• Open access
• Published: 25 April 2011

Malnutrition among children under the age of five in the Democratic Republic of Congo (DRC): does geographic location matter?

• Ngianga-Bakwin Kandala 1 , 2 ,
• Tumwaka P Madungu 3 ,
• Jacques BO Emina 4 ,
• Kikhela PD Nzita 5 &
• Francesco P Cappuccio 6  

BMC Public Health volume  11 , Article number:  261 ( 2011 ) Cite this article

198k Accesses

128 Citations

10 Altmetric

Metrics details

Although there are inequalities in child health and survival in the Democratic Republic of Congo (DRC), the influence of distal determinants such as geographic location on children's nutritional status is still unclear. We investigate the impact of geographic location on child nutritional status by mapping the residual net effect of malnutrition while accounting for important risk factors.

We examine spatial variation in under-five malnutrition with flexible geo-additive semi-parametric mixed model while simultaneously controlling for spatial dependence and possibly nonlinear effects of covariates within a simultaneous, coherent regression framework based on Markov Chain Monte Carlo techniques. Individual data records were constructed for children. Each record represents a child and consists of nutritional status information and a list of covariates. For the 8,992 children born within the last five years before the survey, 3,663 children have information on anthropometric measures.

Our novel empirical approach is able to flexibly determine to what extent the substantial spatial pattern of malnutrition is driven by detectable factors such as socioeconomic factors and can be attributable to unmeasured factors such as conflicts, political, environmental and cultural factors.

Although childhood malnutrition was more pronounced in all provinces of the DRC, after accounting for the location's effects, geographic differences were significant: malnutrition was significantly higher in rural areas compared to urban centres and this difference persisted after multiple adjustments. The findings suggest that models of nutritional intervention must be carefully specified with regard to residential location.

Childhood malnutrition is spatially structured and rates remain very high in the provinces that rely on the mining industry and comparable to the level seen in Eastern provinces under conflicts. Even in provinces such as Bas-Congo that produce foods, childhood malnutrition is higher probably because of the economic decision to sell more than the population consumes. Improving maternal and child nutritional status is a prerequisite for achieving MDG 4, to reduce child mortality rate in the DRC.

Peer Review reports

Malnutrition prevents children from reaching their full physical and mental potential. Health and physical consequences of prolonged states of malnourishment among children are: delay in their physical growth and motor development; lower intellectual quotient (IQ), greater behavioural problems and deficient social skills; susceptibility to contracting diseases [ 1 , 2 ]. Furthermore, child malnutrition is associated with approximately 60 percent of under-five mortality in Sub-Saharan Africa (SSA) countries [ 3 ].

The majority of studies on child nutritional status have described prevalence of malnutrition among under-five children and analyzed socioeconomic, demographic and cultural factors associated with child malnutrition in SSA [ 4 – 7 ]. However, little is known about the links between child's nutritional status and distal determinants including geographic location and the environment due to restricted methodologies.

Our study aims to investigate the impact of geographic location as a proxy for distal factors and their influences on nutritional status of children. The province of residence is taken as a modifiable variable which can help explain the variation of malnutrition among children between different provinces.

Four reasons justify the interest of this study: first, geographic location is an important modifier of known predictors of malnutrition and is associated with food security and accessibility, especially in the context of conflict affected country such as the DRC.

Second, through the use of our empirical methods we can investigate inequalities in childhood malnutrition by mapping the residual net effect of spatial pattern of malnutrition more flexibly than most previous work.

Third, the methods also allow us to investigate non-linear effects of some risk factors prior to and after controlling for the socioeconomic determinants. This enables us to determine to what extent the substantial spatial pattern of malnutrition is driven by socioeconomic factors or point to the influence of omitted variables with strong spatial structure or possibly conflicts, political or environmental and cultural factors or even epidemiological processes that account for this spatial structure.

Fourth, the worsening socioeconomic, cultural and political context of the DRC needs to be investigated. The DRC is one of the SSA countries characterised by extreme poverty, high incidence of childhood diseases, high mortality and poor infrastructure: 75 percent of people are malnourished [ 1 ]; hundreds of thousands of children have died due to malnutrition over the past 12 years [ 3 ]. Furthermore, the country continues to experience armed conflicts and political instability since 1990. However, regardless of the worsening socioeconomic, political and health situations little is known about inequalities in childhood malnutrition across socio-economic strata or provinces although preliminary reports from the existing national surveys highlight the problem of malnutrition among children.

Background on study area

The DRC is the third largest country (by area: 2,344,858 km 2 ) in Africa and with immense natural resources distributed across its 11 provinces. It is, with the population of more than 68 million, the eighteenth most populous nation in the world, and the fourth most populous nation in Africa, 62 percent of which are under the age of fifteen.

Poverty and vulnerability are the main characteristics of the Congolese population. First, the World Bank estimated that the DRC's per capita gross domestic product (GDP) in 1999 was 78 US Dollar. The GDP has since declined. External debt at the end of 2000 was 12.9 billion US$ which, according to the Word Bank, equals roughly 280 percent of the GDP and to 900 percent of the exports. The accumulated debt and severe economic decline are due to both recent war and decades of corruption and economic mismanagement [ 8 , 9 ].

Further, since 1996, the DRC has been hit by conflict, which has devastated and destabilized the country and claimed the lives of an estimated six million civilians [ 10 ]. People continue to live in crisis conditions in many parts of the country. The eastern provinces (Orientale, Katanga, North and South Kivu), and more recently the province of Equateur, are afflicted by violence.

The ongoing Congolese crisis has claimed more lives than any conflict since World War II [ 10 ], and it continues to be of concern to the international community. Despite many political agreements signed since the start of the conflict, there is little expectation and prospect for peace as lives of vulnerable groups such as women and children continue to be shattered as conflict re-emerged in the eastern part of the country and a new front of violence opened in the province of Equateur. These conflicts have continued to hinder the DRC's ability to drive development efforts forward, so the population continues to suffer the consequences. Compounding this situation is the lack of leadership, mismanagement, corruption, rapid deterioration of the socio-economic conditions and the fall of prices of mineral resources which the national economy rely on because of the global financial crisis, which resulted in a sharp drop in revenues and massive loss of employment. Little progress is made in the implementation of the Government's Priority Action Plan on agriculture as most resources are concentrated on the army. Programmes are urgently needed to improve food security and auto-dependence, which would thereby reduce the country's over-reliance on humanitarian interventions to address the long-lasting acute and chronic malnutrition the country, continues to face [ 11 ].

Thus, humanitarian needs in the country remain colossal. According to the Central Emergency Respond Fund report in 2008, conflict has generated up to 1.35 million internally displaced persons (IDPs) in only three provinces, corroding the coping mechanisms of millions of people. With the continuation of conflict and the actions of abusive armed groups have increased food prices, matched with the limited ability of productive areas to feed population centres due to logistic constraints have generated malnutrition rates of up to 20 percent in certain health zones [ 11 ].

Consequently, chronic malnutrition is a serious problem, affecting some 48 percent of children in the DRC [ 12 ].

Preliminary reports from three nutritional national surveys (the 1995 and 2001 Multiple Indicator Cluster Surveys (1995 and 2001 MICS) and the 2007 Demographic and Health Survey (2007 DRCDHS) show that nutritional situation in the DRC remains critical [ 12 , 13 ]. Specifically, nutritional status of children under the age of 5 indicated deterioration in terms of acute malnutrition (stunting, wasting and underweight). Stunting rate was respectively 34 percent in 1995, 31 percent in 2001, and 46 percent in 2007. The nutritional status of mothers is also critical: about 19 percent of them were suffering from low Body Mass Index (BMI) in 2007.

The ever worsening political climate in Eastern provinces, resulted in war since 1996, has created an unprecedented hardship on the population, especially on children as they are more prone than adults to suffer from nutritional deficiencies because of their physiologically less stable situation [ 8 ]. Very high malnutrition rates have been recorded in the war provinces because of insecurity. But even in peace areas untouched by the present conflict nearly half of the children are malnourished [ 14 ]. Malnutrition remains one of the main factors associated with the high childhood morbidity and mortality [ 15 , 16 ].

National estimates of malnutrition may conceal important intra-provincial differences due to diverse cultural norms that might affect nutritional practices and the impact of the ongoing conflict on food security. It is therefore, important to examine patterns of malnutrition at a more disaggregated province level.

We recognise that a province in a country such as the DRC is a large unit of observation, but the provinces' estimates are more informative compared to the use of national estimates of malnutrition.

There is no specific empirical study undertaken to investigate determinants of malnutrition among children in the DRC. We therefore investigate the impact of geographic location on childhood malnutrition while taking into account the effect of the important risk factors of malnutrition present in the DHS database that might confound or mediate the inequalities of the spatial patterns observed at the province-level in order to gain a good understanding of the extent of malnutrition in a post-conflict country. The results will enrich the current literature with recent data on malnutrition, making it more understandable and helping to establish more effective intervention policies to monitor and evaluate achievement of the Millennium Development Goals (MDGs) in countries devastated by conflict. The policy interventions that would not account for unobservable distal factors (such as conflicts, political, environment etc...) will not deliver the required outcomes and will prolong the vulnerability of children in the DRC.

Geographic Location in the DRC DHS

By applying the spatial analysis to the disaggregated province-level, we are able to establish whether the spatial effects cross the boundaries between the provinces or are distinct, which would also give us a sense on the relative importance of policies versus geographic factors in causing malnutrition.

While the eastern provinces used to be the major food producers of the country, repeated looting of crops by armed groups and general insecurity over many years has undermined production.

In other parts of the country with better security conditions, crumbled infrastructure has significantly decreased the country's food production capacity. Households and major food importers maintain food reserves at a bare minimum because of the volatile political and economic environment, as well as the frequent threats of looting.

High prices have also hit the DRC hard. Food prices have increased by 52 percent in June 2009 compared to figures from May 2008 [ 1 ]. This is probably due to the lack of national policy for food production and the reliance of the DRC on food aid (the DRC relies 100% on aid). The financial and economic crisis has also affected mining activities. Acute malnutrition is at dangerously high levels in some parts of the DRC. Acute malnutrition is above the emergency threshold in the Kasaï provinces (centre). Even the worst affected parts of North Kivu do not have such high rates perhaps due to humanitarian interventions. Malaria, malnutrition, acute respiratory infections, tuberculosis, and diarrhea are the main causes of child mortality, according to the Ministry of Health. Deteriorating health conditions have allowed the resurgence of epidemics such as measles and typhoid fever.

As conflict continues to prevail in Province Orientale, South, North Kivu and Equateur, children are subject to starvation, and there is an increase in child mortality and morbidity. An almost total lack of basic health and social infrastructure has had a negative impact on child health.

This study uses data from the 2007 DRC Demographic Health Survey (DRC-DHS), a national representative investigation on children's and women's health. The DRC-DHS data has comparable information on community and household characteristics as well as on nutrition and health of women aged 15-49 years and their children under-five years old at the time of the survey. The samples covered all regions, urban and rural areas. In total 9,000 households (3,690 in urban areas and 5,310 in rural areas) were sampled. All women between the ages of 15 to 49 living in these households were interviewed. Mother and under-five nutritional module covers a sub-sample of one household out of two from the 9,000 selected households. The data contains information on 9,995 women and 8,992 children under the age of 5. The DHS data is of good quality. However, the information provided by this survey is cross-sectional. The samples collected under the DHS survey is drawn together using stratified multistage sampling designs, often with over-sampling of smaller domains such as urban areas or certain regions of a country. In many instances, these data are subsequently analyzed using statistical software designed for simple randomly sampled data. Such analyses fail to take into account the impact of the underlying complex sampling design on regression parameter estimates. Consequently, conclusions drawn from these analyses may give misleading estimates on important health indicators on which public policies are based. Techniques that account for the survey design such as weighting, stratification, and hierarchical regression can be used. Furthermore, DHS data use cluster-sampling to draw upon women respondents via multistage sampling, where: at the first stage, a stratified sample of enumeration areas (villages/communities) is taken; at the second stage, a sample of households within the selected communities is taken; and finally, at the third stage, all women respondents (aged 15-49 years) in the sample households are included. Cluster sampling is a cost-saving measure, without the need to list all the households. However, statistically, it creates analytical problems in that observational units are not independent. Thus, statistical analyses that rely upon the assumption of independence are no longer valid. We focus on the hierarchical regression technique using Bayesian Geo-additive models to take into account the above mentioned issues.

Nutritional status

According to the World Health Organization (WHO) [ 17 ], malnutrition has three commonly used comprehensive types named stunting, wasting and underweight measures by height for age, weight for height and weight for age indexes respectively.

Stunting or growth retardation or chronic protein-energy malnutrition (PEM) is deficiency for calories and protein available to the body tissues and it is inadequate intake of food over a long period of time, or persistent and recurrent ill-health. This height-for-age index (stunting) is less sensitive to temporary food shortages and thus seems to be considered as the most reliable indicator. Because studies have shown that wasting is volatile over seasons and periods of sickness and underweight shows seasonal weight recovery and being overweight for some children can also affect weight-for-age index [ 8 ].

Wasting or acute protein-energy malnutrition captures the failure to receive adequate nutrition during the period immediately before the survey, resulting from recent episodes of illness and diarrhoea in particular or from acute food shortage. Underweight status is a composite of the two preceding ones, and can be due to either chronic, acute malnutrition or PEM.

In the three surveys, nutritional status was assessed according to weight-for-age, weight-for-height and height-for-age using the US National Center for Health Statistics/WHO international reference tables and charts [ 17 , 18 ]. Wasting, stunting and underweight were defined as weight-for-height, height-for-age and weight-for- age of 2SD or more below the corresponding median of the reference population, respectively; while severe wasting, severe stunting was defined as 3SD or more below the same median, respectively.

We focused on stunted children (2 SD of height-for-age below the median value) as our covariates were better able to explain chronic than acute malnutrition. We used the Z-Score (in a standardized form) as a continuous variable to maximize the amount of information available in the data set.

It is worth mentioning that, because of the drawback of the international reference population in correctly capturing nutritional status of children around the world; recently a new reference standard has been generated from which Z-scores can be calculated. For the purpose of this paper, the use of the new reference standard would not change the qualitative results. A detailed discussion on the new reference standard can be found in [ 19 ].

Figure 1 shows a histogram and kernel density estimates of the distribution of the Z-scores, together with a normal density, with mean and variance estimated from the sample. This gave us clear evidence that a Gaussian regression model is a reasonable choice for our inference for the dependent variable stunting.

Histogram, kernel density of stunting (left) and mean standardized Z-score for stunting by child's age (right)

Correlates of Malnutrition

Child nutritional status is actually caused by multiple factors including, but not exclusively, those with illness, disease, and biological causes. A fuller understanding of illness and disease must include considerations of cultural, psychological, social and political factors present in the physical environment where the child lives. This premise has been expanded in many different areas such as medical, child psychology and sociology and now forms a fundamental part of a great deal of social science research and practice.

Mosley and Chen [ 20 ] in their study of the causes of death in children in developing and low income countries, placed risk factors within an analytical framework or including the interactions among socio-economic, cultural, environmental and biomedical factors. The framework focuses on the factors or determinants according to how direct the impact of the determinant was on the risk of death, i.e. the proximity of the risk posed to the children.

The Proximate factors include biological agents of disease e.g., microbes and vectors, and other elements which directly influence child's exposure to the agents of disease and ill health.

Distal factors include features of the wider socio-cultural, environmental and political context affecting both the child; his/her care givers e.g. public health policies and safety as well as cultural norms, environmental degradation which dictate how a family may respond to an illness.

These associations illustrate the vulnerability of children in any population who live in the environment where many of these determinants become unavailable or unstable.

Since we are interested in multiple causes of malnutrition, when modelling the determinants of malnutrition, we can distinguish between immediate, intermediate, and underlying determinants [ 3 ]. While malnutrition is always immediately related to either insufficient nutrient intake or the inability of the body to absorb nutrients (primarily due to illness), these are themselves caused by food security, care practises, and the health environment at the household level, which themselves are influenced by the socioeconomic and demographic situation of households, communities and public health policies [ 3 , 21 , 22 ]. Factors such as food security, care practises and health environment are a matter of public health policies. We refer to them as distal determinants of malnutrition.

In order to capture this complex chain of causation, various approaches have been taken each focusing on a particular level of causality. Studies [ 21 , 23 ] have estimated structural equations that address the interactions; Caputo, et al. [ 24 ] have used graphical chain models to assess the causal pathways, and other studies [ 5 ] have used multi-level modelling techniques. However, with the available data, it is not always clear how to separate intermediate from underlying determinants. For example, mother's education might be influencing care practises, an intermediate determinant, and the resources available to the household, an underlying determinant. On the other hand, child province of residence, a distal determinant, might influence food prices and security, intermediate determinants, and food availability, an underlying determinant.

Given these difficulties, our approach is to estimate models that mainly focus on factors that are mostly underlying determinants of malnutrition, although some might also be considered intermediate determinants and distal determinants. The most important covariate included in this analysis is the geographic location where the child lives that includes features of the wider socio-cultural and political context affecting both the child and his/her care givers. Other selected socio-demographics variables available in the data are grouped as individual child's characteristics, mother's characteristics, household economic level and community's characteristics. Regarding the covariates, we were guided by the previous literature on the subject and the conceptual framework outlined in [ 3 ].

Unfortunately, the surveys do not generate an income variable and we therefore rely on a household asset index as a proxy for the socio-economic status of the households which has been found to be quite reliable. Ownership of consumer items, such as a radio or car, as well as characteristics of the dwelling such as floor or roof type, toilet facilities and water source are items that measure poverty in these setting and the World Bank and others have used these items to generate an asset index, using Principal Components Analysis (PCA). We use the first principal component derived from the data to obtain the index for each household. We sort children by the asset index and establish cut-off values for percentiles of the population. We then refer to the bottom third as 'low socioeconomic status, the next third as 'medium socioeconomic status, the top third as 'high socioeconomic status' (see Table 1 ).

Among the underlying determinants of chronic malnutrition, we considered as a proxy measure of current or recent socioeconomic status (SES), the asset index, household size, the nutritional status of the mother (measured by her BMI), health knowledge and care practices measured by mother's education, mother's marital status, birth interval and place of delivery of children.

We also control for the sex of the child, urban rural location, and the age of child. Based on prior own work as well as other literature [ 17 , 22 , 23 ], we investigated a potentially non-linear pattern of effects of the mother's BMI as well as the age pattern on malnutrition. For illustration, the empirical distribution of the stunting Z-score by child's age is shown in Figures 1 (right). It is obvious that the effect of child's age on the mean Z-score of stunting is nonlinear. It will be difficult to model the possibly nonlinear effect of such covariates through a parametric functional form, which well justifies our use of a flexible semi-parametric model. Empirical distributions of all factors used in the analysis, are given in Table 1 .

Statistical analysis

Historically, variations in malnutrition prevalence has been related to household socio-economic factors because it determines the amount of resources (such as food, good sanitation, and health care) that are available to infants and neglected temporal and geographic gradients and other variations in risk, in order to generate hypotheses towards the cause of malnutrition.

We examine spatial variation in under-five malnutrition with flexible geo-additive semi-parametric mixed model while simultaneously controlling for spatial dependence and possibly nonlinear effects of covariates within a simultaneous, coherent regression framework. Individual data records were constructed for children. Each record represents a child and consists of nutritional status and a list of covariates. For the 8,992 children born within the last five years before the survey, 3,663 children have information on anthropometric measures. Because the predictor contains usual linear terms, nonlinear effects of metrical covariates and geographic effects in additive form, such models are also called geo-additive models. Kammann [ 25 ] proposed this type of models within an empirical Bayesian approach. Here, we apply a fully Bayesian approach as suggested in [ 26 ] which is based on Markov priors and uses Markov Chain Monte Carlo (MCMC) techniques for inference and model checking. For model choice, we routinely used the Deviance Information Criterion (DIC) developed in Spiegelhalter et al. [ 27 ], as a measure of fit and model complexity.

Geo-additive and geo-referenced disaggregated province level or site-specific analysis is a means of managing spatial and temporal variability of determinant of different types: distal, proximate and intermediate factors which are deemed to affect child nutritional status.

The aim of site-specific province analysis is to accelerate policy interventions, optimise inputs (unobserved factors such as distal ones: food security and prices policies, environmental etc...), improve child nutrition by taking into account the environmental impact and reduce the timescale to attain the Millennium Development Goals (MDGs). It is an approach that deals with multiple groups of factors input to improve child nutritional status in order to satisfy the actual needs of parts of the provinces rather than average needs of the whole country.

The analysis was carried out using version 0.9 of the BayesX software package [ 28 ], which permits Bayesian inference based on MCMC simulation techniques. The statistical significance of apparent associations between potential risk factors and stunting was explored in chi-square and Mann-Whitney U -tests, as appropriate. Multivariate analysis was used to evaluate the significance of the posterior mean determined for the fixed, non-linear effects and spatial effects. A P -value of < 0.05 was considered indicative of a statistically significant difference. We also run a sensitivity analysis for the choice of priors. Standard choices for the hyper-parameters are a = 1 and b = 0:005 or a = b = 0:001: Je?rey's Non-informative prior is closer to the later choice, and since practical experience shows that regression parameters depend on the choice of hyper-parameters, we have investigated in our application the sensitivity to this choice.

It would be beyond the scope of this paper to go into the details of estimation procedures. Please refer to Appendix 1 for a detailed explanation of the statistical methods. The method has also been discussed in more detail in [ 22 ].

Table 1 shows individual characteristics of the sample population prior to multiple adjustments of all factors that might confound or mediate the observed spatial variation within provinces on stunting.

Of the overall sample of 8,992 children, 41 percent (3,663) of the sample children had measurement on their height and weight to ascertain their nutritional status. Of those 50.8 percent was female and the overall prevalence of malnutrition (stunting) was 43.9 percent.

The prevalence of stunting was higher among boys compared to girls (46.1 versus 41.7 percent), has an inverse linear association with the age of the child (higher in the age groups ranging from 4 years, followed by 3 years, 2 years, 1 years but lower in the younger age (0 year): 55.1, 49.4, 48.5, 46.5 versus 23.1 percent), higher in rural areas compared with urban areas (48.4 versus 37.2 percent), higher among children born outside the hospital compared with their counterpart born in hospitals (49.1 versus 41.8 percent), linearly associated with maternal education (higher among children from non educated mother, followed by children from mothers with primary education but lower among children from mothers with secondary or higher education: 49.8, 47.0 versus 35.2 percent ), linearly associated with socio-economic status of the household (higher among children from the poorest household, followed by children from poor, middle or rich households but lower among children from richest households: 49.8, 48.0, 45.5, 43.9 versus 28.7 percent ), very high in Sud Kivu (46.1 percent) and Kasai Occidental (46.1 percent) provinces, followed by Nord Kivu (45.0 percent), Katanga (44.4 percent), Bandundu (42.4 percent), Kasai Oriental (42.0 percent), Bas Congo (40.3 percent), Maniema (39.1 percent), Equateur (36.7 percent), Orientale (35.3 percent) provinces, but lower in Kinshasa, the capital city (16.4 percent).

On the other hand, there were no statistically significant association observed between the prevalence of stunting and gender of the household's head, mother's marital status, preceding birth interval of the child, and household's size.

The geographical distribution of the crude prevalence of the standardized Z-scores for the response variable stunting by province display in Table 1 shows distinct spatial patterns. While in Kinshasa, Orientale and Equateur provinces, it appears that stunting was lower, there seem to be more areas of high stunting in North-Eastern of the DRC that is affected by conflict and the three provinces that relied heavily on local mineral mining (Katanga and the two Kasai). In addition to local small-area variability, there might also be an underlying smooth spatial component, which crosses provincial borders due to displacement of population during the conflicts, something we investigated below. The provincial prevalence shown in Table 1 also suggested that we should examine the spatial pattern of stunting at a more disaggregated province level as the national prevalence of 43.9 percent glossed over important intra-province differentials.

In the multivariate analysis the results for the fixed effects in Table 2 suggest that female children are slightly less stunted, as found in other studies [ 22 , 29 ]. In fact, the corresponding posterior mean, -0.12 for male, is negative and the 10% and 90% quintiles are both negative - indicating that the effect is statistically significant. Children living in rural areas are more stunted than their counterpart in urban areas. Maternal education rather than paternal education has a positive impact on children's nutritional status as well as household's socio-economic status. Children from low socioeconomic households were, as expected, more stunted than children from high income backgrounds.

We also estimated the posterior mean of stunting and plotted it against child's age and mother's BMI. As hypothesised, Figure 2 shows that there is a bell shaped, non-linear relationship between the effects of child's age (left), mother's BMI (right) and stunting. Shown are the posterior means together with the 80% and 95% pointwise credible intervals. As found in other countries of SSA [ 22 ], these data show that the effect of mother's BMI on child's nutritional status to be in the form of an inverse U shape. While the inverse U looks nearly symmetric, the descending portion exhibits a much larger range in the credible region. This appears quite reasonable as obesity of the mother (possibly due to a poor quality diet) is likely to pose less of a risk for the nutritional status of the child as very low BMIs, which suggest acute undernutrition of the mother [ 22 ]. The Z-score is highest (and thus stunting lowest) at a BMI of around 30-35. The figure also shows that there are few women with high BMI (40 or higher) in the survey, but this is likely to represent an artefact of the small numbers sampled at this BMI range.

Non-linear effects of and child's age (left) and mother's body mass index (right) on stunting . Shown are posterior mean of stunting within the 80% and 95% credible interval

Figure 2 left shows the effect of the child's age on its nutritional status. As hypothesised and commonly suggested by the nutritional literature [ 22 ], we are able to discern the continuous worsening of the nutritional status up until about 20 months of age. This deterioration sets in right after birth and continues, more or less linearly, until 20 months. Such an immediate deterioration in nutritional status is not as expected as the literature typically suggests that the worsening is associated with weaning at around 4-6 months. One reason for this finding could be that, according to the surveys, most parents give their children liquids other than breast milk shortly after birth, which might contribute to infections at these early ages.

After 20 months, stunting stabilizes at a low level. Through reduced growth and the waning impact of infections, children are apparently able to reach a low-level equilibrium that allows their nutritional status to stabilize.

We also see a sudden improvement of the Z-score around 24 months of age. This is picking up the effect of a change in the data set that makes up the reference standard. Until 24 months, the currently used international reference standard is based on white children in the US of high socioeconomic status, while after 24 months; it is based on a representative sample of all US children [ 17 ]. Since the latter sample exhibits worse nutritional status, comparing the Congolese children to that sample leads to a sudden improvement of their nutritional status at 24 months [ 17 , 22 ].

This anomaly of the reference standard is one reason for the replacement of this reference population by a new reference standard from the WHO [ 19 , 29 ].

Figure 3 explores province specific net spatial effects of undernutrition. We report results of the model that includes the total residual spatial effects of the province (i.e. the sum of both the structured and unstructured spatial effects). The left panel of Figure 3 shows the total residual spatial effects of the province and the right panel of Figure 3 indicates the significance of the observed spatial effects in the form of a posterior probability map. The levels correspond to significantly negative (black colour), significantly positive (white colour) and insignificant (grey colour). Three important observations emerge. First, there is a strong north-south gradient in these provincial effects with a fairly sharp dividing line running through the centre of the country. Over and above the impact of the fixed effects, there appear to be negative influences of malnutrition in the south-east that are quite general and affect most of the provinces there. Given that the south-eastern provinces are all affected by the ongoing conflict than the rest of the country, it is likely that food security and price policies, environmental factors and associated conflict e.g. relying on food aids and, lack of public infrastructure, lack of farming due to conflicts are responsible for this pronounced regional pattern. Therefore, humanitarian assistance that the population mostly relies on in these conflict-affected provinces might have short-term impact on child nutritional status. Second, living in the capital Kinshasa and Sud-Kivu is associated with significantly better nutrition despite Sud Kivu being affected by the conflict and surrounded by provinces with negative effects (Nord Kivu and Maniema). Note that both rates of prevalence of stunting in Kinshasa and Sud-Kivu are above the emergency threshold of 15 percent. As in most developing countries, living in the capital provides access to nutrition and health care that is superior in ways that have not been captured adequately in the fixed effects. The advantage in nutritional status of children living in Sud- Kivu may be due to the fact that the province receives more food aid than any other province in the DRC. Many aids organizations are based in this province and there has been an influx of food aids in this province. Therefore, in the province of Sud-Kivu, children have probably more benefited from international food assistance. In other provinces that are affected by conflicts such as Nord-Kivu where many aid organizations are also based particularly in Goma and there has been an influx of food aid in this province, it is surprising that many children still suffer from severe malnutrition even though food is abundant where they live. One possible explanation is that the lack of food is due to the fear of cultivation in unsecured environment. Another possible explanation is that most children in these provinces live in displacement camps and the higher intensity of the conflict due to the predation of the abundant mineral resources in this area by armed groups [ 15 ].

Total residual spatial effect of stunting (left) and posterior probabilities (right) of stunting for the full model

In the two Kasai , one explanation for the high rate of stunting may be due to the fact that the first livelihood activities are wage labour and mining activities and few people are involved in agriculture. This explanation might also be true for for higher malnutrition observed in Katanga, which also relies on mining in addition to the impact of war.

To compare our province-specific nonlinear spatial effects with our simple fixed effects for provinces (Table 1 ), Figure 3 presents a map that shows those provincial effects. One can only distinguish three main provinces effects. Better nutritional status is found in the Orientale province and Equateur province as well as Kinshasa, worse nutritional status in the eastern provinces under conflicts and non significant effect for provinces in the south of the DRC. In contrast, the crude provincial fixed effects shown in Table 1 miss most of the findings we discussed above. In particular, the sharp North-South gradient present in the province analysis is clearly now visible as the three eastern provinces include provinces on both sides of that divide. Moreover, the positive effect of Kinshasa is simply averaged in with the Bas Congo and Bandundu provinces. Clearly, a lot is lost when relying on these crude estimates of modelling spatial effects.

The DHS data provides a consistent, large and national database that can be used to analyze patterns of malnutrition in the DRC. This study has shown the relationships between malnutrition and the geographic location as well as a number of other risk factors that could explain the site-specific variation at the province level.

Our results show that children's chronic malnutrition is highly prevalent in the entire country with rates largely above 40 percent. The DRC has a deficit of food and limited food productivity despite the country's enormous potential for agricultural production. Only the western part of the country is a net producer, in particular the province of Bas Congo.

Over the last ten years, there has been a significant decline of the production of almost all agricultural products. According to the World Food Programme (WFP), the production of cassava has decreased by 23 percent between 1992 and 2006; the production of plantain has decreased by 75 percent between 1990 and 2006. There has been an increase of the maize production (by 33 percent between 1990 and 2006) however in Maniema and North Kivu the production has decreased by 22, in Katanga by 12 percent.

The deterioration in food productivity is the result of many factors but can be attributed mainly to distal factors such as lack of implementation of national policy for food production, security and conflicts. The agricultural system is mainly subsistence-oriented. According to the WFP, more than 93 percent of households have access to land, however the majority cultivates less than 1 hectare, which does not allow for adequate production for sale or own consumption. Cultivation techniques are still very traditional and households lack farming tools. Few households have a plough or a tractor. Agricultural inputs, such as fertilizers are not available. Eight years after the launch of the government PMURR programme (Programme Multi sectoriel des Urgences pour la Reconstruction et la rehabilitation) to make fertilizers available to farmers, the programme has yet to make an impact on the agricultural sector. Also, the year 2010 was declared by the government as the agricultural year to push many reforms in the sector but the impact of such programmes is yet to be seen.

Seeds are often of low quality, and productivity is low. These are clearly areas where if there were a national policy, this could make a difference for the DRC. Also, in the Eastern provinces people do not cultivate due to the violence, in the provinces such as Katanga, the two Kasai and Orientale, the young generation has left the agricultural sector to work in the mining industries (gold, diamond and coltan). In the eastern provinces, only 18 percent of households own livestock. When they do, it is usually in small quantity. Goat is the main livestock owned [ 30 ].

The results of these rates are similar to the one of the other countries [ 30 ]. Likewise, the risk of stunting is higher in rural areas, among children from less educated mothers and living in poorer household after controlling for other variables in the model [ 7 ]. As in most countries of SSA, there is substantial spatial province difference in child nutritional status in the DRC. Kinshasa's population is essentially urban, the proportion of the most educated women is higher compared to other provinces and accessibility to health facilities and safe drinking water is better whereas rural children, or less educated mothers have difficult access to health facilities, and consume about half the calories daily than their urban counterparts [ 1 ].

The major finding of this study is that malnutrition rates remain very high in the provinces that rely on the mining industry (two Kasai and Katanga) comparable to the level seen in Eastern provinces under war. One possible explanation may be found in the nutritional behaviour of the population that do not give certain types of food to children on cultural grounds even though the food is nutritious and in the reliance of the population living in these provinces on artisanal mining industry and the neglect of agriculture. A survey on food security showed that Kasai occidental has the worst indicator on population availability for food. There is a real hunger problem in this province, because the population that lives in this province does not want to work in agriculture and it prefers to work in the traditional extraction of diamonds. Even in provinces such as Bas-Congo that produces foods, the population sells more than it consumes. The higher rate of malnutrition observed in the eastern provinces under war is not surprising; the lack of food is due to insecurity rather than their inability to produce food because these provinces are known as traditionally pastoral and agricultural provinces.

Another observation drawn from this paper (Table 1 ) is the gap in malnutrition rates between the province of Kinshasa and all other provinces. In fact, Kinshasa's stunting prevalence is very low compared with the national rate. But it is above the emergency threshold by humanitarian standard. In spite of the generalized state of poverty in the country, incomes are higher in Kinshasa; as a result, economically, the population of Kinshasa enjoys better access to food products. The presence of more educated mothers and their partners in Kinshasa, and the lowest rate of poorest people living there may permit better nutritional practices.

The strong evidence of statistically significant difference of malnutrition between socio economic groups mainly between poorest, poorer, middle and richer groups compared to the richest group confirms the reality that in the DRC affording food for the majority of the population is still a challenge [ 1 ]. According to the WFP, about 55 percent of households' expenditure is spent on food (only 40 percent in Bandundu). The main source of food is people's own production. The second source of food is the market, except for the two provinces of Kivu, where households rely first on the markets to access food.

Hence, in richer households, often children are well fed and cared for and provided with a safe and stimulating environment, through which they are more likely to survive, to have fewer diseases and illnesses, and to fully develop thinking, language, emotional and social skills [ 12 ]. But in poorer households, most children are affected by the resurgence of kwashiorkor - lack of proteins in the diet - although this remains controversial. This is certainly due to the increasing poverty among parents who cannot afford to buy proteins (groundnuts, beans, meat, fish, and milk) for their children. Findings are largely consistent with findings of others studies on malnutrition by socio economic status (SES) in SSA [ 7 ] and highlight that poorer children have a higher risk of becoming stunted than richer ones.

The gap observed on stunting prevalence between children from uneducated mothers or those whose mothers have a primary school level of education compared with those from mothers with secondary or high level of education remains high. In fact, education could make a difference by empowering mothers (decision on type of nutrition and/or use of preventive medicine). Similar results have been found in Cameroon [ 7 ] and in most developing countries [ 21 ]. Education could also help the mothers make informed nutritional decisions about cultural norms on certain types of food for children.

With reference to other variables, male children seem to be more exposed to the risk of malnutrition than female children. There is no obvious explanation for this gender difference but in Asia, for instance, gender's difference has been attributed to boys' preference over girls [ 29 ]. Also, older children are more prone to be exposed to anthropometric failure than their counterparts aged less than one. Mainly, older children are mixed breastfed, even not breastfed at times, while younger children may be protected by the mother's immune system at birth [ 22 ]. The risk could be also due to lack of foods in the households due to poverty or the lack of hygiene by mothers, when cooking children foods.

The direct causes of malnutrition are the lack of access to drinking water (in the DRC, it is estimated that more than two thirds of the population has no access to drinking water), morbidity (malaria, respiratory infections and diarrhoea) and poor food consumption [ 22 , 30 ]. Also, breast feeding practices are inadequate and according to the WFP, about 12 percent of the under 18 children are orphans. The prevalence changes significantly across the country, and it is higher in the East (more than 16 percent in province Orientale) [ 30 ].

This study has been able to determine that in the DRC, childhood malnutrition is spatially structured and rates remain very high in the provinces that rely on the mining industry and comparable to the level seen in Eastern provinces under war. In war-affected provinces, we are able to determine that childhood malnutrition is higher probably because of the environmental impact caused by war because these provinces are known as traditionally pastoral and agricultural provinces. Furthermore, the massive influx of population especially from Rwanda, Uganda and Sudan fleeing conflicts has further exacerbated the food crisis. Food aids has helped but it is unsustainable. Even in provinces such as Bas-Congo that produce foods, childhood malnutrition is higher because of the economic decision to sell more than the population consumes.

In summary, in the DRC the improvement of the nutritional status of children would help avert child deaths from diarrhoea, pneumonia, malaria, HIV and measles. Consequently it would reduce neonatal mortality, helping achieve MDG 1, which main aim is to reduce poverty and hunger. There is an urgent need for national policies to improve the security of people and implement agricultural policies for auto-dependent agriculture (the DRC has the potential with plenty of land for agriculture). In other words, improving maternal and child nutrition is a prerequisite for achieving MDG 4, to reduce the child mortality rate. Also, nutritional programmes and policies that will try to reduce female illiteracy and provide basic infrastructures in rural areas in order to reduce gaps in health care between socio-economic groups are likely to succeed. The majority of the poorest household lives in rural areas and poorest children are more exposed to the risk of being malnourished. Hence, there is an urgent need to build programmes which aim to reduce poverty in both rural and urban areas, and which will take into account inequalities observed between provinces in the DRC.

Classical linear regression models of the form

for observations ( y i , w i ), i = 1,...., n , on a response variable y and a vector w of covariates assume that the mean E ( y i | w i ) can be modeled through a linear predictor w i ' γ . In our application to childhood under-nutrition and in many other regression situations, we are facing the following problems: First, for the continuous covariates in the data set, the assumption of a strictly linear effect on the response y may not be appropriate. In our study, such covariates are the child's age ( age ), the mother's age at birth ( mab ), and the mother's Body Mass Index ( BMI ). Generally, it will be difficult to model the possibly nonlinear effect of such covariates through a parametric functional form, which has to be linear in the parameters, prior to any data analysis.

Second, in addition to usual covariates, geographical small-area information was given in form of a location variable s , indicating the province, district or community where individuals or units in the sample size live or come from. In our study, this geographical information is given by the provinces of the DRC. Attempts to include such small-area information using province-specific dummy-variables would in our case entail more than 50 dummy-variables and using this approach we would not assess spatial inter-dependence. The latter problem cannot also be resolved through conventional multilevel modeling using uncorrelated random effects. It is reasonable to assume that areas close to each other are more similar than areas far apart, so that spatially correlated random effects are required.

To overcome these difficulties, we replace the strictly linear predictor through a geo-additive predictor , leading to the geo-additive regression model

here, f 1 ,...,f p are non-linear smooth effects of the metrical covariates, and f spat is the effect of the spatial covariate s i ∈ {1,..., S } labelling the provinces in the DRC. Regression models with predictors as in (2) are sometimes referred to as geo-additive models. In a further step we may split up the spatial effect f spat into a spatially correlated (structured) and an uncorrelated (unstructured) effect: f spat (s i ) = f str (s i ) + f unstr (s i ) . The rationale is that a spatial effect is usually a surrogate of many unobserved influences, some of them may obey a strong spatial structure and others may be present only locally. The observation model (2) may be extended by including interaction f(x)w between a continuous covariate x and a binary component of w , say, leading to so called varying coefficient models, or by adding a nonlinear interaction f 1,2 (x 1 , x 2 ) of two continuous covariates.

In a Bayesian approach unknown functions f j and parameters γ as well as the variance parameter σ 2 are considered as random variables and have to be supplemented with appropriate prior assumptions. In the absence of any prior knowledge we assume independent diffuse priors γ j α const, j = 1,...,r for the parameters of fixed effects. Another common choice is highly dispersed Gaussian priors.

Several alternatives are available as smoothness priors for the unknown functions f j ( x j ), see [ 26 ]. We use Bayesian P(enalized) - Splines,. It is assumed that an unknown smooth function f j (x j ) can be approximated by a polynomial spline of low degree. The usual choices are cubic splines, which are twice continuously differentiable piecewise cubic polynomials defined for a grid of k equally spaced knot p on the relevant interval [ a , b ] of the x-axis. Such a spline can be written in terms of a linear combination B-spline basis functions B m ( x ), i.e.

These basis functions have finite support on four neighbouring intervals of the grid, and are zero elsewhere. A comparably small number of knots (usually between 10 and 40) is chosen to ensure enough flexibility in combination with a roughness penalty based on second order difference of adjacent B-spline coefficients to guarantee sufficient smoothness of the fitted curves. In our Bayesian approach this corresponds to second order random walks

with Gaussian errors u m ~ N (0,τ 2 ). The variance parameter τ 2 controls the amount of smoothness, and is also estimated from the data. More details on Bayesian P-Splines can be found in [ 28 ]. Note that random walks are the special case of B-Splines of degree zero.

We now turn our attention to the spatial effects f str and f unstr . For the spatially correlated effect f str (s), s = 1, ... S, we choose Markov random field priors common in spatial statistics. These priors reflect spatial neighbourhood relationships. For geographical data one usually assumes that two sites or regions s and r are neighbours if they share a common boundary. Then a spatial extension of random walk models leads to the conditional, spatially autoregressive specification

where N s is the number of adjacent regions, and r ∈ ∂ s denotes that region r is a neighbour of region s. Thus the (conditional) mean of f str (s) is an average of function evaluations f str (s) of neighbouring regions. Again the variance τ 2 str controls the degree of smoothness.

For a spatially uncorrelated (unstructured) effect f unstr a common assumption is that the parameters f unstr (s) are i.i.d. Gaussian

Variance or smoothness parameters τ 2 j , j = 1,..., p , str , unstr , are also considered as unknown and estimated simultaneously with corresponding unknown functions f j . Therefore, hyper-priors are assigned to them in a second stage of the hierarchy by highly dispersed inverse gamma distributions p(τ 2 j ) ~ IG(a j , b j ) with known hyper-parameters a j and b j . For model choice, we routinely used the Deviance Information Criterion (DIC) developed in [ 27 ], as a measure of fit and model complexity.

FAO: State of Food Insecurity in the World. 2008, Retrieved on 30 March, 2010, [ http://www.fao.org/es/ess/faostat/foodsecurity/index_en.htm ] : Food Security Statistics

Google Scholar  

Black R, Morris S, Jennifer B: Where and Why Are 10 Million Children Dying Every Year?. The Lancet. 2003, 361: 2226-2234. 10.1016/S0140-6736(03)13779-8.

Article   Google Scholar  

UNICEF: The state of the world's children 1998. 1998, New York, UNICEF

Reed BA, Habicht JP, Niameogo C: The effects of maternal Education on child nutritional status depend on socio-environmental conditions. International Journal of Epidemiology. 1996, 25: 585-592. 10.1093/ije/25.3.585.

Article   CAS   PubMed   Google Scholar  

Madise NJ, Matthews Z, Margetts B: Heterogeneity of child nutritional status between households: a comparison of six sub-saharan African countries. Population Studies. 1999, 53: 331-343.

Gwatkin DR, Rustein S, Johnson S, Pande K, Wagstaff A: Socioeconomic differences in Health, Nutrition and population in Cameroon. 2000, HNP/Poverty Thematic Group of World Bank, Washington DC

Pongou R, Ezzati M, Salomon JA: Household and Community Socioeconomic and Environmental Determinants of Child Nutritional Status in Cameroon. BMC, Public Heath. 2006, 6.98: 19-

World Bank: World Development Indicators 2009. 2009, Washington DC: the World Bank

Book   Google Scholar  

UNDP: D.R. Congo Human Development Report. 2007, New York, NY: United Nations Development Programme

Coghlan B, Brennan RJ, Ngoy P, Dofara D, Otto B, Clements M, Stewart T: Mortality in the Democratic Republic of Congo: a nationwide survey. The Lancet. 2006, 367: 44-51. 10.1016/S0140-6736(06)67923-3.

Central Emergency Respond Fund (CERF): The Democratic Republic of Congo. Facts and Figures. 2008, Retrieved 30 March, 2010, [ http://ochaonline.un.org/cerf/CERFHome/tabid/1705/language/en-US/Default.aspx ]

UNICEF: The state of the world's children 2007. 2007, New York, UNICEF

République Démocratique du Congo: (RDC), Enquête démographique et de Santé (EDS-RDC 2007). Rapport final. 2008

United States Agency for International Development: The Democratic Republic of Congo. Retrieved on 30 March, 2010, [ http://www.usaid.gov/pubs/cbj2003/afr/cd/ ]

Doctor without Borders: Food, nutrition and mortality situation of IDP's in Dubie. Katanga 23-25 March 2006

Kandala NB, Emina JBO, Nzita DK, Cappuccio FP: Diarrhoea, acute respiratory infection, and fever among children in the Democratic Republic of Congo. Social Science and Medicine. 2009, 68: 1728-1736. 10.1016/j.socscimed.2009.02.004.

Article   PubMed   Google Scholar  

World Health Organization (WHO): WHO Technical Report Series No. 854. Physical Status: The Use and Interpretation of Anthropometry. 1995, Geneva: WHO

World Health Organization (WHO): Measuring Change in Nutritional Status. 1983, Geneva

World Health Organization (WHO): Multicentre Growth Reference Study. Acta Paediatrica. 2006, 450 (Suppl): 1-87.

Mosley W, Chen L: An Analytical Framework for the Study of Child Survival in Developing Countries. Population and Development Review. 1984, 10: 25-45.

Smith L, Haddad L: IFPRI Research Report No. 111. Explaining Child Malnutrition in Developing Countries. 1999, Washington DC: IFPRI

Kandala NB, Fahrmeir L, Klasen S, Priebe J: Geo-additive models of Childhood Undernutrition in three Sub-Saharan African Countries. Population, Space and Place. 2009, 15: 461-473. 10.1002/psp.524.

Moradi A, Klasen S: SFB 386 Discussion Paper No. 217. The Nutritional Status of Elites in India, Kenya, and Zambia: An Appropriate Guide for Developing International Reference Standards for Undernutrition?. 2000, University of Munich

Caputo A, Roraita R, Klasen S, Pigeot I: Undernutrition in Benin: An. Analysis based on Graphical Models. Social Sciences and Medecine. 2003, 56: 1677-1691. 10.1016/S0277-9536(02)00162-4.

Kammann EE, Wand MP: Geoadditive Models. Journal of the Royal Statistical Society C. 2003, 52: 1-18.

Fahrmeir L, Lang S: Bayesian Inference for Generalized Additive Mixed Models Based on Markov Random Field Priors. Applied Statistics (JRSS C). 2001, 50: 201-220.

Spiegelhalter D, Best N, Carlin B, Van der Line A: Bayesian measures of models complexity and fit. Journal of the Royal Stat Soc B. 2002, 64: 1-34. 10.1111/1467-9868.02022.

Brezger A, Kneib T, Lang S: BayesX: Analysing Bayesian structured additive regression models. Journal of Statistical Software. 2005, 14 (11):

Klasen S: Poverty, undernutrition, and child mortality: Some inter-regional puzzles and their implications for research and policy. Journal of Economic Inequality. 2008, 6 (1): 89-115. 10.1007/s10888-007-9056-x.

United Nations World Food Programme (WFP): Democratic Republic of Congo Comprehensive Food Security and Vulnerability Analysis (CFSVA) / 2007-2008. 2008

Pre-publication history

The pre-publication history for this paper can be accessed here: http://www.biomedcentral.com/1471-2458/11/261/prepub

Download references

Acknowledgements

This research was supported by the British Council under the DelPHE (Development Partnership in Higher Education) scheme. The authors thank Macro international, for providing free the 2007 DHS data-sets for the DR Congo.

Author information

Authors and affiliations.

Warwick Medical School, Health Sciences Research Institute, University of Warwick, Warwick Evidence, Gibbet Hill, CV4 7AL,, Coventry, UK

Ngianga-Bakwin Kandala

Department of Population Studies, University of Botswana, Gaborone, Private box 0075, Botswana

Institut National de Statistique, Kinshasa, Congo

Tumwaka P Madungu

African Populations and Health Research Center, Shelter Afrique Centre, Longonot Road, P.O.Box 10787, 00100 GP.O, Nairobi, Kenya

Jacques BO Emina

Département des Sciences de la Population et du Développement, Faculté des Sciences Economiques, Université de Kinshasa, B.P. 176, Kinshasa XI, Congo

Kikhela PD Nzita

University of Warwick, Warwick Medical School, Clinical Sciences Research Institute, Clifford Road Bridge, CV2 2DX, Coventry, UK

Francesco P Cappuccio

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Ngianga-Bakwin Kandala .

Additional information

Competing interests.

The authors declare that they have no competing interests.

Authors' contributions

N-BK: Conception and design, literature review, data analysis and interpretation, drafting the article, critical revisions for important intellectual content and approval of final article for submission;

PTM: Literature review, interpretation of results, drafting the article, critical revisions for important intellectual content and approval of final article for submission;

JBOE: Interpretation of results and critical revisions for important intellectual content;

PDNK: Interpretation of results and critical revisions for important intellectual content;

FPC: Interpretation of results and critical revisions for important intellectual content; and all authors read and approved the final manuscript.

Authors’ original submitted files for images

Below are the links to the authors’ original submitted files for images.

Authors’ original file for figure 1

Authors’ original file for figure 2, authors’ original file for figure 3, rights and permissions.

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( http://creativecommons.org/licenses/by/2.0 ), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Reprints and permissions

About this article

Cite this article.

Kandala, NB., Madungu, T.P., Emina, J.B. et al. Malnutrition among children under the age of five in the Democratic Republic of Congo (DRC): does geographic location matter?. BMC Public Health 11 , 261 (2011). https://doi.org/10.1186/1471-2458-11-261

Download citation

Received : 26 April 2010

Accepted : 25 April 2011

Published : 25 April 2011

DOI : https://doi.org/10.1186/1471-2458-11-261

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Gross Domestic Product
• Spatial Effect
• Eastern Province
• Acute Malnutrition
• Asset Index

BMC Public Health

ISSN: 1471-2458

• Submission enquiries: [email protected]
• General enquiries: [email protected]

• Bahasa Indonesia

Water security is critical for poverty reduction, but billions will remain without water access unless urgent action is taken

WASHINGTON, May 20, 2024 — Access to safe drinking water and sanitation, reliable water-supply for agriculture and industry, and protection against droughts and floods are essential for human and economic development, a World Bank report released on Monday states.

Over the past 20 years, the number of people lacking safe drinking water and basic sanitation has increased by 197 million and 211 million, respectively. Today, over two billion people still lack access to safe drinking water, and 3.5 billion are deprived of safely managed sanitation facilities. Resulting infectious diseases contribute to at least 1.4 million annual deaths and 50% of global malnutrition.

Lack of access to safe water and sanitation is particularly harmful in childhood, Water for Shared Prosperity , a report released at the 10 th World Water Forum in Bali, Indonesia by the World Bank Group and the Government of Indonesia, says. Inadequate and unsafe water affects early childhood development, and time spent fetching water, inadequate sanitation and hygiene and droughts or floods disrupt learning and lead to school dropouts.

Climate change is amplifying water-related risks. Driven by global emissions, developing countries are most affected by climate shocks. Between 2000 and 2021, developing countries experienced more severe droughts and longer lasting floods than advanced economies, with long-term effects on nutrition, school attendance and economic welfare. Developing countries disproportionately rely on water-dependent sectors, particularly agriculture, for employment. Globally, over 800 million people are at high-risk of droughts and twice as many live in flood-risk hotspots.

“To improve livelihoods, significant reforms and investments are needed to provide efficiently managed water and sanitation services to those without access, and to strengthen resilience against hydro-climatic risks,” said World Bank Vice President for East Asia and the Pacific Manuela V. Ferro , who is leading the World Bank team at the World Water Forum.

Water for Shared Prosperity offers specific recommendations on how to improve water security in developing countries: Protecting depleting aquifers and unevenly distributed freshwater resources will require more international cooperation, implementing proven nature-based solutions such as reforestation and investing in water storage infrastructure to prevent run-off and make water available in dry periods.

Policies to upgrade housing, and land use regulations to prevent construction in flood-prone areas can reduce exposure. Early warning systems and insurance can help households and farmers cope with extreme hydro-climatic shocks.

Reforming water tariffs and poorly-targeted subsidies, while ensuring affordability for low-income households, can help maintain and expand services and allocate scarce water resources fairly. Service providers will also need to enhance their operations, reducing water losses and lowering operating costs. Supported by policies that ensure transparency and accountability, the private sector can offer valuable expertise to enhance efficiency and manage complex infrastructure.

A spotlight section of the report examines how Indonesia, the host of this year's triennial World Water Forum, is addressing water security challenges. Indonesia has made significant investments to enhance resilience to climate-related risks, including investments in 61 dams to store water and increase irrigated areas.  A Community-Based Water Supply Program has provided more than 24 million people with improved water facilities. The government has prioritized reducing pollution and environmental degradation in the Citarum River Basin in West Java, and is pioneering treatment of peat water to making it suitable for drinking through the National Urban Water Supply project.

This site uses cookies to optimize functionality and give you the best possible experience. If you continue to navigate this website beyond this page, cookies will be placed on your browser. To learn more about cookies, click here .

The Federal Register

The daily journal of the united states government, request access.

Due to aggressive automated scraping of FederalRegister.gov and eCFR.gov, programmatic access to these sites is limited to access to our extensive developer APIs.

If you are human user receiving this message, we can add your IP address to a set of IPs that can access FederalRegister.gov & eCFR.gov; complete the CAPTCHA (bot test) below and click "Request Access". This process will be necessary for each IP address you wish to access the site from, requests are valid for approximately one quarter (three months) after which the process may need to be repeated.

An official website of the United States government.

If you want to request a wider IP range, first request access for your current IP, and then use the "Site Feedback" button found in the lower left-hand side to make the request.