Determinants of anemia in school-going adolescents: a case study in Douala, Cameroon | BMC Public Health

Determinants of anemia in school-going adolescents: a case study in Douala, Cameroon | BMC Public Health

Setting and study period

The study was conducted in 4 secondary schools (Bonamoussadi Polyvalent Private Institute, Douala Bilingual High School, PK 21 High School, and Douala Bassa Technical High School) selected in four arrondissements of the city of Douala (Douala 2nd, Douala 3rd, Douala 4th, and Douala 5th). in the city of Douala. It extended over six months from July to December 2023. The city of Douala is a cosmopolitan city with approximately 4.9 million inhabitants spread over an area of 41,000 ha [24]. It is one of the two largest cities in the country, with the political capital Yaoundé. The city grew from its commercial port on the estuary of the Wouri River, which opened up to the Gulf of Guinea. It is the capital of the Littoral region and the Wouri department and is an urban community composed of six municipalities [25].

This site was selected because of the lack of data on anemia in adolescents in this region. Despite the size and importance of Douala, there is a lack of specific research concerning the risk factors and prevalence of anemia in this age group. Thus, this study aims to fill this gap by providing crucial information that can help guide public health policies and targeted interventions to improve the health of young people in this area.

Study design and population

The survey methodology used in this study was that used by Gonete et al., in 2018 [26]. The study was quantitative and cross-sectional with descriptive and analytical purposes. The overall population of the study consisted of all adolescents of both sexes residing permanently in the city of Douala. The target population consisted of all adolescents aged 10 to 15 attending pre-selected schools. Parental consents were sent to parents via the children, and only those whose parents had signed the informed consent and returned it to the school were included in the study. Exclusion criteria included adolescents with known chronic illnesses, those on medication that could affect study results, and those who were unable to provide valid parental consent.

Sample size and procedure

The sample size was calculated using Lorentz’s formula: n = z2 × p × (1-p) / e2, where n is the sample size, z the confidence level (1.96 for a 95% confidence level), p the estimated 32% prevalence of anemia in adolescents, taken from a previous study [27]. Considering a planned effect of 1.3 and adding 10% for possible non-response, this study included 511 adolescent women..

The sampling technique used for this study was non-probabilistic and based on convenience. Four secondary schools (one private, the Institut privé polyvalent de Bonamoussadi, and three public, namely the Lycée bilingue de Douala, the Lycée PK 21 and the Lycée technique de Douala Bassa) were selected in four arrondissements of the city of Douala (Douala 2nd, Douala 3rd, Douala 4th and Douala 5th). This selection was made to avoid the concentration of samples in one homogeneous area, which could distort the results. By including a variety of schools, we ensure that the demographic, cultural and economic characteristics of the adolescents are taken into account, thus reducing the risk of selection bias. Thus, by selecting one school per district, we have ensured geographical and socio-economic diversity, contributing to more representative results.

Data collection technique and tools

The data collection technique was the survey in the school environment. The data were collected using a structured questionnaire. This enabled us to collect data on socio-demographic variables, dietary diversity, menstrual history, iron supplementation, hemoglobin levels, and anthropometric parameters. The questionnaire was first pre-tested with 10 adolescents aged between 10 and 15, after which their understanding of the questions and the length of the interview were assessed. The questionnaire was administered in English and French, depending on the schools surveyed.

To determine the level of hemoglobin (Hb), 10 μL of capillary blood samples were taken by tapping the tip of the finger with a single-use sterile lancet. The first drop of blood was wiped and the second drop was used to determine Hb. The analysis of hemoglobin was carried out using a hemoglobinometer in which, after blood migration within this device, the measurement was made electronically [28].

About the anthropometric parameters, the weight was measured at almost 0.1 kg using a CAMRY® brand weight with a capacity of 160 kg without shoes and socks with as few clothes as possible, the blister empty, and the pockets of the clothing pre-empt. The height was measured at 0.1 cm nearby using a Stadiometer (height chart). The teenager stood upright, the horizontal indicator of the Stadiometer was lowered until it tightly touched the crown of the head, and the sliding plate of the towel allowed them to locate the upper benchmark [29].

The hemoglobinometer was regularly calibrated in accordance with the manufacturer’s specifications, ensuring accurate measurements of hemoglobin levels. Similarly, the CAMRY® balance and stadiometer were checked for accuracy prior to the start of the study. Regular calibration and maintenance practices were followed to ensure that all instruments provided consistent and reliable measurements throughout the data collection process. This attention to the reliability of measurement tools reinforces the validity of the study results.

Data collection procedure

Data collection was carefully planned to ensure compliance with ethical protocols and adolescent participation. One week before the start of data collection, consent forms were sent to the students’ parents via the adolescents themselves. Only those whose parents signed and returned the consent forms were included in the study. This method made families aware of the importance of the study and ensured informed participation. Once consent had been obtained, data were collected through a structured questionnaire administered in the selected schools. The collection of blood samples to measure hemoglobin levels and anthropometric measurements (weight and height) were also carried out as part of this procedure, ensuring a comprehensive and systematic approach to the assessment of anemia in adolescents.

Operational definition of variables

Anemia and its severity were defined based on the World Health Organization recommended threshold rates for non-pregnant adolescents: (Hb > 11 g/dL) no anemia, (Hb 11 to 11.9 g/dL) mild anemia, (Hb 8 to 10.9 g/dL) moderate and (Hb < 8 g/dL) severe anemia. The Body Mass Index (BMI) was calculated by dividing the weight in kilograms by the square of the height in meters (kg/m2). The BMI benchmarks used to diagnose malnutrition (< 18.4 kg/m2), normal state (18.5 and 24.9 kg/ m2), overweight (25–29.9 kg/m2), and obesity (≥ 30 kg / m2) were those recommended by the World Health Organization [4].

The level of dietary diversity was assessed on a scale of 0 to 3, based on the frequency of consumption of three food groups: fruit, vegetables and meat/fish, at mealtimes (breakfast, lunch or dinner) in the week preceding the study. The classification was established as follows: individuals consuming none of the three foods were given a very low level (0), those consuming one of the three foods were given a low level (1), those consuming two foods were classified with a medium level (2), and those consuming all three foods were given a high level (3).

To assess physical activity, participants were asked about the frequency and duration of their sporting activities over the course of a week (running, walking, soccer, volleyball, etc.). This variable was coded into two categories: “Yes” for those who regularly take part in physical activity (defined by a recommended minimum duration, e.g. 30 minutes a day) and “No” for those who do not.

Iron supplementation was assessed via a questionnaire in which participants indicated whether they had taken iron supplements (tablets, syrup, etc.) in the last three months. Responses were coded as “Yes” for those who had used supplements and “No” for those who had not.

This variable was determined by asking participants if they had contacted any of these diseases (malaria, sickle cell disease, viral hepatitis, amoebiasis, typhoid fever, etc.) in the past 3 months. Responses were coded as “Yes” for those who had suffered from an illness (with examples such as respiratory infections, gastroenteritis, etc.) and “No” for those who had not.

The variable concerning parental car ownership was used as an indicator of family wealth. Respondents were classified as “Yes” if their parents owned a car, and “No” if they did not.

The study variables fall into two broad types. Dependent variables include anemia and its severity, measured by Hemoglobin (Hb) levels. Independent variables include Body Mass Index (BMI), used to assess nutritional status, as well as the frequency of consumption of different food groups (fruit, vegetables, meat/fish) during meals, which may influence the onset of anemia.

Data processing and analysis

The data collected through the Open Data Kit Collect application were compiled and analyzed using the Rstudio analytics software version 4.2.4. The Microsoft Office Excel 2013 software was used for the elaboration of the tables. Descriptive statistics were calculated for all variables considered in this study. The association between the dependent variable “anemia status” and the independent variables was assessed using binary logistic regression analysis. Logistic regression assumptions include independence of observations, linearity of the relationship between continuous independent variables and the log odds of the dependent variable, and absence of multicollinearity. To ensure that these assumptions were met, we carried out specific checks on linearity using graphical analyses such as scatter plots, and assessed multicollinearity using Variance Inflation Factor (VIF) values. This approach enabled us to confirm that our model meets the assumptions necessary to obtain reliable results. All variables with a P value < 0.05 in bivariate analysis were candidates for multivariate analysis, to control for confounders and identify independent predictors of anemia. Missing data were treated by imputation to minimize their impact on the results. The various indicators were estimated with an alpha (α) = 0.05 degree of significance and the associations were considered to be significant for a P < 0.05.

link

Leave a Reply

Your email address will not be published. Required fields are marked *