Soil-transmitted helminth infection and nutritional status among urban slum children in Kenya

To evaluate the nutritional impact of soil-transmitted helminth (STH) infection, we conducted a cross-sectional survey of 205 pre-school (PSC) and 487 school-aged children (SAC) randomly selected from the surveillance registry of the Centers for Disease Control and Prevention of the Kibera slum in Kenya. Hemoglobin, iron deficiency (ID), vitamin A deficiency (VAD), inflammation, malaria, anthropometry, and STH ova were measured. Poisson regression models evaluated associations between STH and malnutrition outcomes and controlled for confounders. Approximately 40% of PSC and SAC had STH infection, primarily Ascaris and Trichuris; 2.9% of PSC and 1.1% of SAC had high-intensity infection. Malnutrition prevalence among PSC and SAC was anemia (38.3% and 14.0%, respectively), ID (23.0% and 5.0%, respectively), VAD (16.9% and 4.5%, respectively), and stunting (29.7% and 16.9%, respectively). In multivariate analysis, STH in PSC was associated with VAD (prevalence ratio [PR] = 2.2, 95% confidence interval = 1.1-4.6) and ID (PR = 3.3, 95% confidence interval = 1.6-6.6) but not anemia or stunting. No associations were significant in SAC. Integrated deworming and micronutrient supplementation strategies should be evaluated in this population. Copyright © 2014 by The American Society of Tropical Medicine and Hygiene. The study was conducted in Kibera in southern Nairobi, Kenya, which is one of the largest contiguous urban slums in Africa.10 The Centers for Disease Control and Prevention’s International Emerging Infections Program (IEIP), in collaboration with the Kenya Medical Research Institute (KEMRI), has conducted community-based morbidity surveillance in Kibera since 2005.11 IEIP performs active household surveillance for major infectious disease syndromes for approximately 28,000 participants living in 2 of Kibera’s 13 villages. Field workers make home visits every 2 weeks to ask about recent illnesses; they also maintain an up-to-date registry by enrolling new arrivals and updating records to reflect outmigration.10,11 Kibera is characterized by high population density, semipermanent housing, and lack of official city water or sewage services. Participants can access free healthcare at a centrally located clinic staffed by study-supported trained personnel and other clinics run by non-governmental organizations and government institutions bordering the slum. Among households in the IEIP participant registry, approximately 25% of households were designated as potential sources for enrollment of pre-school children (PSC; ages 6–59 months), and the other 75% of households were designated as potential sources for enrollment of school-aged children (SAC; ages 5–14 years); this designation ensured that PSC and SAC were not from the same household. Households were then selected with probability proportional to size from each target group, and one child was randomly chosen from each selected household. Sampling weights were calculated to account for non-response and post-stratified to the IEIP cluster size populations. Sample size was capped by the laboratory capacity for daily processing of fresh stools for STH. At 80% power and accounting for 20% non-response, the achievable target sample size of 293 PSC and 899 SAC was powered to detect an odds ratio of 2.0 for STH infection associated with having an infected sibling. Written informed consent was obtained from all participating households. Subjects with anemia, positive malaria test, wasting, or STH infection were referred for free care at the nearby IEIP-run clinic. Institutional review boards of KEMRI and the Centers for Disease Control and Prevention (CDC) approved this study. Fieldwork took place from April to June of 2012. Trained fieldworkers administered a mobile device-based questionnaire to obtain demographic and socioeconomic data and child breastfeeding, deworming, and nutrition supplementation history. Anthropometric measurements of height and length were taken using a wooden measuring board accurate to 0.1 cm (Irwin Shorr Productions, Olney, MD). Weight was measured to the nearest 0.1 kg using a digital scale (Seca Corp, Hanover, MD). Trained fieldworkers completed the measurements using standard techniques. Capillary blood samples were collected for hemoglobin (Hb) measurements, malaria testing, and analysis of micronutrient and inflammatory status. A goal of three stool samples was collected from selected children. Stools were collected over 3 consecutive days when possible, accepted only if produced after midnight the previous night, and maintained in cool boxes before delivery to the laboratory for analysis before 14:00 hours each day. Stool samples were processed at KEMRI using the Kato–Katz technique. Two slides were prepared per sample; approximately 7% of all slides were reread for quality control. Details of the nutrition laboratory analysis are described elsewhere.12 In brief, Hb was determined using a HemoCue Hb301 machine (Ängelholm, Sweden). Anemia was defined by age and corrected for altitude according to WHO guidelines13: Hb < 11.0 g/dL in children 6–59 months of age, Hb < 11.5 g/dL in children 5–11 years of age, and Hb < 12 g/dL in children 12–14 years of age; 0.6 g/dL was added to cutoffs to account for an altitude of 1,800 m. A rapid diagnostic test kit (SD BIOLINE Malaria Ag P.f/Pan, Hagal-dong, Korea) was used to test for malaria. Blood samples were centrifuged, and frozen plasma samples were transported to VitA-Iron Tech (Willstaett, Germany), where levels of ferritin, retinol binding protein (RBP), C-reactive protein (CRP), and α-1-acid glycoprotein (AGP) were measured by sandwich enzyme-linked immunosorbent assay (ELISA).14 The following thresholds were used to define abnormal values for these biochemical indicators: ferritin < 12 μg/L in PSC and ferritin < 15 μg/L in SAC15; RBP < 0.7 μmol/L; CRP > 5 mg/L; AGP > 1 g/L.16 Because biomarkers of nutrition are influenced by inflammation, iron deficiency, vitamin A deficiency, and anemia were defined using a correction factor approach to adjust ferritin, RBP, and Hb values for the presence of inflammation, which was described previously.16,17 Because the prevalence of inflammation varied by age, separate correction factors were calculated for PSC and SAC. Among PSC, correction factors for ferritin, RBP, and Hb were as follows: 0.87, 1.0, and 0.91 for early inflammation (elevated CRP and normal AGP); 0.36, 1.51, and 1.08 for early convalescent inflammation (elevated CRP and AGP); 0.78, 1.1, and 1.0 for late convalescent inflammation (elevated AGP and normal CRP). Among SAC, correction factors for ferritin, RBP, and Hb were as follows: 0.70, 1.36, and 1.01 for early inflammation; 0.66, 1.36, and 1.09 for early convalescent inflammation; 0.99, 1.21, and 1.08 for late convalescent inflammation. We used the WHO Child Growth Standards (WHO Anthro, Geneva, Switzerland) to calculate z-scores and categorized underweight as a weight-for-age z-score of < −2, stunting as a height/length-for-age z-score of < −2, and wasting as a body mass index (BMI)-for-age z-score of < −2 to allow for comparisons between PSC and SAC. We used the weekly expenditure on food by the household to classify respondents by socioeconomic status. Children were classified as having received or not received vitamin A supplementation in the past 1 year based on written documentation when available or otherwise, respondent recollection. Statistical analyses were performed using R version 2.15.2 (R Core Team, 2012). The survey package was used to incorporate survey weights. Comparisons between PSC and SAC were performed with χ2 test. Means and standard deviations (SDs) were computed for continuous variables, and t test was used to compare means. Multivariable analyses used Poisson regression to provide prevalence ratios (PR).18 All analyses were run separately for PSC and SAC. Five nutritional outcomes were investigated: anemia, vitamin A deficiency (VAD), iron deficiency (ID), iron deficiency anemia (IDA), and stunting. Inflammation was also investigated as an outcome (e.g., elevated CRP and/or elevated AGP). These outcomes were dichotomized according to previously described cutoffs. Several alternative measures of the primary exposure (STH infection) were used: presence/absence of any STH infection; presence/absence of species-specific infection with Ascaris or Trichuris; presence/absence of coinfection with both Ascaris and Trichuris; and intensity of infection (none, light, medium/moderate, or high according to standard egg per 1 gram-based WHO cutoffs19 using the highest-intensity species found in the case of coinfection, with eggs per 1 gram determined as an average among all the subjects' stools). Demographic and clinical confounders were determined independently for each outcome a priori based on knowledge from published literature that they were potential predictors of the various nutritional outcomes. Demographic factors considered were age, sex, maternal education, ethnicity, socioeconomic status, and breastfeeding. Clinical factors included were malaria, stunting, wasting, ID, and VAD. To test the robustness of the model, we also performed a sensitivity analysis for the VAD outcome, which included history of recent vitamin A supplementation. Unadjusted prevalence ratios and 95% confidence intervals were computed for association between a specific outcome and primary exposure. Age, sex, and socioeconomic status were forced in the models as potential confounders. When the primary exposure was Ascaris infection, presence of Trichuris infection was also forced in the models. The decision to include the other demographic variables and clinical variables in the models was based on at least a 10% change in the prevalence ratio of the primary exposure. Variables that were originally continuous were treated as both continuous and dichotomized variables in separate models. The results of these models were compared to assess residual confounding and whether associations were linear. The best models selected were those models with confidence intervals that were more precise.