An evaluation of an operations research project to reduce childhood stunting in a food-insecure area in Ethiopia.

OBJECTIVE: To determine which interventions can reduce linear growth retardation (stunting) in children aged 6-36 months over a 5-year period in a food-insecure population in Ethiopia. DESIGN: We used data collected through an operations research project run by Save the Children UK: the Child Caring Practices (CCP) project. Eleven neighbouring villages were purposefully selected to receive one of four interventions: (i) health; (iii) nutrition education; (iii) water, sanitation and hygiene (WASH); or (iv) integrated comprising all interventions. A comparison group of three villages did not receive any interventions. Cross-sectional surveys were conducted at baseline (2004) and for impact evaluation (2009) using the same quantitative and qualitative tools. The primary outcome was stunted growth in children aged 6-36 months measured as height (or length)-for-age Z-scores (mean and prevalence). Secondary outcomes were knowledge of health seeking, infant and young child feeding and preventive practices. SETTING: Amhara, Ethiopia. SUBJECTS: Children aged 6-36 months. RESULTS: The WASH intervention group was the only group to show a significant increase in mean height-for-age Z-score (+0·33, P = 0·02), with a 12·1 % decrease in the prevalence of stunting, compared with the baseline group. This group also showed significant improvements in mothers' knowledge of causes of diarrhoea and hygiene practices. The other intervention groups saw non-significant impacts for childhood stunting but improvements in knowledge relating to specific intervention education messages given. CONCLUSIONS: The study suggests that an improvement in hygiene practices had a significant impact on stunting levels. However, there may be alternative explanations for this and further evidence is required.

The effect of body shape on weight-for-height and mid-upper arm circumference based case definitions of acute malnutrition in Ethiopian children.

BACKGROUND: Nutritional anthropometry surveys from Somalia and Ethiopia have reported that standard weight-for-height z-score (WHZ) and mid-upper arm circumference (MUAC) case definitions return different estimates of the prevalence of acute malnutrition in pastoralist livelihood zones but similar estimates of the prevalence of acute malnutrition in the agrarian livelihood zones. A study undertaken in Somalia to investigate this finding reported that children from pastoralist livelihood zones tended to have longer limbs and lower SSRs than children from agrarian livelihood zones. AIM: The present study investigated the relationship between weight-for-height and body shape and the relationship between MUAC and body shape in different populations of Ethiopian children. SUBJECTS AND METHODS: Six cross-sectional nutritional anthropometry surveys were undertaken. The combined survey datasets form the study sample. Data sources were grouped according to the livelihood zone from which data originated (either settled agrarian or semi-nomadic pastoralist). Case definitions of acute malnutrition using WHZ calculated using the NCHS and WHO reference populations and MUAC uncorrected for age or height were used. The SSR was used as an index of body shape. The association between body shape and the different case definitions of acute malnutrition were investigated using standard statistical techniques. RESULTS: Weight-for-height and MUAC case definitions yielded similar estimates of the prevalence of acute malnutrition in agrarian children but different estimates of the prevalence of acute malnutrition in pastoralist children. These populations also exhibit different SSRs. The SSR is an important predictor of weight-for-height. The SSR is a poor predictor of MUAC. CONCLUSION: WHZ and WHZ case status in children are associated with body shape and may overestimate the prevalence of acute malnutrition in some populations. Consideration should be given as to whether WHZ should be replaced by MUAC for the purposes of estimating the prevalence of acute malnutrition.

Anaemia, iron status and vitamin A deficiency among adolescent refugees in Kenya and Nepal.

OBJECTIVE: To investigate the prevalence of anaemia (haemoglobin<11.0 to 13.0 g dl(-1) depending on age and sex group), iron deficiency (transferrin receptor concentration>8.3 microg ml(-1)) and vitamin A deficiency (serum retinol <0.7 micromol l(-1)) in adolescent refugees. DESIGN: Cross-sectional surveys. SETTING: Kakuma refugee camp in Kenya and seven refugee camps in Nepal. SUBJECTS: Adolescent refugee residents in these camps. RESULTS: Anaemia was present in 46% (95% confidence interval (CI): 42-51) of adolescents in Kenya and in 24% (95% CI: 20-28) of adolescents in Nepal. The sensitivity of palmar pallor in detecting anaemia was 21%. In addition, 43% (95% CI: 36-50) and 53% (95% CI: 46-61) of adolescents in Kenya and Nepal, respectively, had iron deficiency. In both surveys, anaemia occurred more commonly among adolescents with iron deficiency. Vitamin A deficiency was found in 15% (95% CI: 10-20) of adolescents in Kenya and 30% (95% CI: 24-37) of adolescents in Nepal. Night blindness was not more common in adolescents with vitamin A deficiency than in those without vitamin A deficiency. In Kenya, one of the seven adolescents with Bitot's spots had vitamin A deficiency. CONCLUSIONS: Anaemia, iron deficiency and vitamin A deficiency are common among adolescents in refugee populations. Such adolescents need to increase intakes of these nutrients; however, the lack of routine access makes programmes targeting adolescents difficult. Adolescent refugees should be considered for assessment along with other at-risk groups in displaced populations.

An evaluation of the impact of a US$60 million nutrition programme in Bangladesh.

OBJECTIVE: To compare levels of childhood malnutrition in areas where the Bangladesh Integrated Nutrition Project had been operational for over 5 years with matched non-project areas, with the purpose of evaluating whether the project had achieved its objective of reducing the prevalence of underweight among children <24 months. METHODS: The study involved an ex-post cross-sectional survey in six thanas (a locality with a population of approximately 200,000-450,000 people) in Bangladesh. Participants were 6,820 households (4,554 in the project areas and 2,266 in the non-project areas) including 7183 children aged 6-59 months selected using a two-stage stratified cluster sampling frame. Main outcome measures were moderate and severe underweight, wasting and stunting reported using z scores, and indicators of mothers' reported nutritional knowledge and practice. RESULTS: 2,388 children aged 6-23 months and 6815 children aged 6-59 months had clean anthropometric data. No significant difference was found between the socio-economic variables of households in the project and non-project areas. No significant difference was found in the prevalence of either severe or moderate underweight (weight-for-age) in children aged 6-23 months in the project and non-project areas: 183 (11.4%, 95% confidence interval 9.9-13.2%) children in project areas and 96 (12.2%, 95% confidence interval 9.9-14.8%) children in non-project areas. Mothers in project areas reported significantly better caring practices than in non-project areas. CONCLUSION: There is no evidence that the Bangladesh Integrated Nutrition Project has achieved its objectives to reduce severe underweight by 40% if project areas are compared ex-post with non-project areas. There is urgent need to review the evidence behind investments based on growth monitoring and promotion.