Menu
Menu
Menu
Menu
Background: Evidence on the differential impacts of the global food crisis as it translates into chronic food insecurity locally is essential to design food security interventions targeting the most vulnerable population groups. There are no studies on the extent of chronic food insecurity or its predictors among adolescents in developing countries. In the context of increased food prices in Ethiopia, we hypothesized that adolescents in low income urban households are more likely to suffer from chronic food insecurity than those in the rural areas who may have direct access to agricultural products. Methods: This report is based on data from the first two rounds of the Jimma Longitudinal Family Survey of Youth (JLFSY). Both adolescents and households were selected using a stratified random sampling method. A total of 1911 adolescents aged 13-17 years were interviewed on their personal experiences of food insecurity both at baseline and at year two. Multivariable logistic regression analyses were used to compare chronic adolescent food insecurity by household income, household food insecurity, and socio-demographic variables after one year of follow-up. Results: Overall, 20.5% of adolescents were food insecure in the first round survey, while the proportion of adolescents with food insecurity increased to 48.4% one year later. During the one year follow up period, more than half (54.8%) of the youth encountered transient food insecurity that is, either during the first or the second round survey. During the follow up period, 14.0% of adolescents had chronic food insecurity (i.e. were food insecure at both rounds). Multivariable logistic regression analysis showed that adolescents in the urban households with low (OR = 1.69, P = 0.008) and middle (OR = 1.80, P = 0.003) income tertiles were nearly twice as likely to suffer from chronic food insecurity compared with those in high income tertile, while this was not the case in rural and semi-urban households. Female sex of adolescents (P < 0.01), high dependency ratio (P < 0.05) and household food insecurity (P < 0.001) were independent predictors of chronic adolescent food insecurity in urban, semi-urban, and rural areas, while educational status of the adolescents was negatively associated with chronic food insecurity (OR = 0.047, P = 0.002) in urban areas. © 2012 Belachew et al.
This study made use of data from 1911 adolescents enrolled in the first two consecutive years of the five year longitudinal study of adolescents in Jimma zone in Southwest Ethiopia. Adolescents were included in the study from urban (Jimma city), semi-urban (small towns) and six rural communities (“Kebeles”) adjacent to the small towns. Details of the methods are presented elsewhere [19]. In brief, from 3700 households that were randomly selected, one male or female adolescent of 13-17 years old was selected from each household using a Kish table [20] with a targeted sample size of 2100 youths (700 each from urban, semi-urban and rural areas). These age groups were selected in order to capture life events as adolescents progressed to adulthood over the course of the longitudinal study. The questionnaire was pilot tested before each round of the survey with 200 adolescents in Jimma city who were not included in the main study, and modified accordingly. Interviews were conducted by full-time employees of the study who received an intensive one-week training prior to the pre-test and an additional training for one week before beginning of the actual interviews, followed by periodic training sessions during the course of the study. Supervisors followed the field procedures closely and checked the completed questionnaires every day to ensure accuracy of the data. The research team made weekly supervisory visits to the whole field team. Ethical approval was obtained from Ethical Review Boards of both Brown University (USA) and Jimma University (Ethiopia). Both heads of households and adolescents gave informed verbal consent before interviews in both rounds of the survey. The first round data were collected from mid 2005 to 2006, while the second round were finalized during the same period between 2006 to 2007. Both household and adolescent questionnaires were interviewer-administered. The household questionnaire gathered information on monthly income, residence, household food security, dependency ratio and gender of the household head. The head of the household responded to the household questionnaire. The adolescent questionnaire gathered information on age, sex, educational status and adolescent food security. The questionnaires were translated prior to the interviews and their consistency was checked by a person who spoke both languages. Interviews were conducted in either of the two local languages (Amharic or Oromifa). Household food insecurity was measured using a six item scale adapted from scales validated for use in developing countries [21-23]. Adolescent food insecurity was measured using a four-item scale adapted by selecting four items from the six-item household food security scale that apply to adolescents’ personal experiences as described elsewhere [24]. The two food security items that were dropped were the ones asking about other children in the household, and adolescents were not therefore expected to answer them. Adolescents were instructed to think of their own experiences, not those of the household, in answering the questions. To avoid bias in the responses related to food security, adolescents and household heads were interviewed separately. For each round of the survey, respondents who answered “yes” to any of these questions were coded as experiencing food insecurity in that survey round. Chronic adolescent food insecurity was defined here as the proportion of adolescents who responded affirmatively to at least one of the four food security item in both rounds of the survey, one year apart. Similarly, chronic household food insecurity is the proportion of households who responded affirmatively to at least one of the six food insecurity items in both rounds of the survey [9]. We used background characteristics from the first round of the survey including household income, dependency ratio, age, highest school grade completed by the adolescent, maternal education, paternal education and sex of the household head as predictor variables. Specifically, we expected that adolescents that were part of households with higher income, with lower proportions of dependent people, with higher education level or those who are male would have better access to food and be less likely to have chronic food insecurity. Household income was divided into tertiles coded as “high”, “middle” and “low”. Household dependency ratio was calculated based on age classifications as the ratio of people who are potentially expected to be nonproductive (age groups greater than 64 and less than 15 years) to people who are expected to be potentially productive (age 15-64 years). After calculating dependency ratio for each household, it was rank ordered and divided into tertiles of “high”, “middle”, or “low” dependency ratio. Buffering was defined as protection of adolescents from chronic food insecurity by adult household members within chronically food insecure households. In this case, buffering is the proportion food secure adolescents within chronically food insecure households. The data were entered in double, checked for missing values and outliers, and analyzed using SPSS (SPSS Inc. version 16.1, Chicago, Illinois). First, bivariate analyses were carried out to identify candidate variables for the multivariable model. Means and proportions were compared using t-test and Pearson’s Chi-square tests after checking all model assumptions. Second, to identify the predictors of chronic food insecurity, only variables that were significantly associated with chronic food insecurity in the bivariate models were entered in the multivariable logistic model. At this step, interaction between different variables was checked. Third, since we expected income to have a different relationship between food security in the urban cash-based economy and in rural agricultural areas with more direct access to food items, and as there was an interaction between place of residence and household income (Pinteraction = 0.01), we stratified the analysis by place of residence and fitted the model for rural, semi urban and urban areas separately. For all scenarios, chronic adolescent food insecurity was used as the dependent variable. Covariates included sex, household income tertile, sex of head of the household, dependency ratio, sex of the head of the household and the highest school grade completed by the adolescent. Covariates were entered with a forward procedure. All tests were two-sided and P < 0.05 was considered statistically significant. We report the results as Odds Ratios (OR) and 95 percent Confidence Intervals (CI). Finally, to assess the resilience of buffering within the households, in relation to household income, we did a cross tabulation of chronic adolescent food insecurity by household income stratified by place of residence for adolescents in chronically food insecure households.
N/A
