Predictors of chronic food insecurity among adolescents in Southwest Ethiopia: a longitudinal study

As a consequence of the recent rise in food prices [1–3], the proportion of people experiencing food insecurity has increased dramatically [3]. Although food prices have fallen since their peak in mid-2008, they still remain high by historic standards [4]. The food price increases in Ethiopia were larger than the increase in world markets and those observed in most other African countries [5, 6]. In Ethiopia, from 2007 to 2008 there was a 100% increase in wheat and teff prices; and maize and meat prices increased by 180% and 50%, respectively [1]. The increase in food prices has led to a reduction in the quality and the quantity of food consumed, particularly amongst vulnerable households who spend a large share of their income on food [1, 7, 8]. Protracted lack of adequate access to food among these vulnerable groups induces chronic food insecurity [9]. Unlike transient food insecurity, which is temporary shortage of food, chronic food insecurity is a long-term challenge, associated with poverty and lower socioeconomic status [6, 10, 11].

Chronically food insecure individuals are exposed to long-term inadequate diets caused by the inability to acquire food [12]. When food prices increase, households that do not produce food themselves [e.g. many urban households] must acquire food in other ways, largely by purchasing food. As the ability to purchase food depends on household income, the rise in the food price can negatively impact on the purchasing power of households [1, 6, 12–14]. Therefore, strategies to cope with food insecurity include reduction of meal frequency and quantity, resorting to less diversified and lower quality cheaper foods [15, 16]. As a result of this prolonged limited access to nutrients, chronic food insecurity could have more deleterious consequences compared to transient food shocks. Moreover the fact that adolescents are in a period of rapid growth and development with an increased demand for energy and other nutrients, and that they do not have decision making power on their access to food within the household makes them more vulnerable to the consequences of chronic food insecurity.

Since Ethiopia’s growth and development strategy identifies food security as a key issue for sustained economic growth and poverty reduction, achieving food security is high on the agenda of the Ethiopian government [17]. Thus, evidence on the differential effects of the global food crisis as it locally translates into chronic food insecurity is essential to design interventions that target the most vulnerable groups. Transient food insecurity is a common phenomenon in Ethiopia [7, 8, 16, 18]. Previous research on adolescents in southwest Ethiopia [16, 18] reported the prevalence of transient food insecurity and how adult household members buffered adolescents from its consequences [16]. However, the extent of chronic food insecurity and how adults protect adolescents against food insecurity when it becomes chronic and household income declines has not been investigated. We hypothesized that adolescents who are members of urban low income households are more likely to suffer from chronic food insecurity. This study intended to answer questions: Does the relation between household income and chronic food insecurity vary by place of residence? How resilient is intra-household buffering of adolescents from food insecurity when purchasing power is dwindling and food insecurity becomes chronic? Does this relationship vary by place of residence?

Overall, 1911 adolescents responded to both the first and the second round food security survey, while 173 were lost to follow-up. The response rate was 91.7%. There was no significant difference in food insecurity at baseline (first round) survey between adolescents who participated in the second round survey and those who dropped out.

In the first round of the survey, 20.5% of adolescents had transient food insecurity, while the proportion of adolescents with transient food insecurity rose to 48.4% one year later. In general, more than half (54.8%) of the adolescents encountered transient food shocks either during the first or the second round survey. During the follow up period, 14.0% of adolescents were chronically food insecure, i.e., they were food insecure in both survey rounds (Figure 1).

The results of bivariate analyses showed that chronic food insecurity was associated with the background characteristics in the first round survey. The proportion of adolescents with chronic food insecurity increased as income tertiles decreased (P value for trend <0.001). Adolescents who were members of households with high dependency ratio had a higher frequency of chronic food insecurity compared to those with low dependency ratio (P = 0.002). The proportion of female adolescents (15.9%) who were chronically food insecure was significantly (P < 0.001) higher than that of males (11.1%). A higher proportion (22.6%) of adolescents who were members of food insecure households in the first round of the survey were chronically food insecure compared to 8.3% of those in food secure households (P < 0.001). Similarly, adolescents with educational status of primary level were twice as likely to be chronically food insecure compared to those with an educational status of secondary and above (14.4% vs. 7.0%, P < 0.001).

The findings of this study show that urban adolescents in low income households are more at risk of suffering from chronic food insecurity than semi-urban and rural adolescents, ceteris paribus. We also found that chronic food insecurity is much higher among adolescents who are members of households in the lower income tertile due to low purchasing power which is highly affected by food price fluctuation. Chronic food insecurity is a problem associated with poverty, affecting social groups with the weakest and most fragile access to food, both in terms of access to social networks and safety nets or productive assets. People’s access to food depends both on the purchasing power of their income and on their non-market mechanisms such as rights to land for subsistence farming and foraging purposes. When the income level is low, households resort to a number of coping strategies to gain access to food and protect food security levels [1, 7, 8, 15]. The coping strategies used could be food stretching (reducing the size of meal), food rationing (reducing the number of times eaten), eating low quality or cheap foods and getting food in socially stigmatized ways such as begging [15].

High food prices further erode the coping capacities of many households across the developing world. When food prices rise, populations at risk of chronic food insecurity are urban low income households and rural landless and pastoralist households which do not have direct access to food items but use money to buy food [7, 8, 25]. The complex relationship between the availability of financial resources in households that access food through purchasing and their food security has been well described [12, 26, 27]. Families depending on monthly income often have a cyclic pattern of food restrictions, with food shortages most severe at the end of the month when household resources are depleted [12, 27, 28]. Chronic food insecurity can be linked to poverty in a bi-directional way that can often lead to a vicious circle. Both past and present economic disadvantage are associated with food insecurity [14], and poor households are often the most vulnerable to chronic food insecurity. Chronic food insecurity in addition, deepens poverty, not just in terms of negative nutritional effects on health and livelihoods, but through the use of irreversible coping mechanisms such as sale of productive assets which make it harder for families to escape from the poverty trap [29]. There is a need for social protection that underpins both production and consumption [30] to address the global commitments to Millennium Development Goals of decreasing the prevalence of poverty and hunger. In the study area, urban poor households need to be targeted through food security interventions as they are the primary victims of the consequences of the rise in the food prices. In Ethiopia food prices have increased considerably following the global food crisis. To respond to the consequences of rapid increase in food prices, the Ethiopian government has initiated a market stabilization program since 2007 [25]. Our results imply that the effectiveness of such intervention would be maximized if the urban poor are targeted.

In food constrained circumstances adults decrease their food intake to protect their children from food insecurity [16, 31–33]. However, this might not guarantee that children’s food intakes are unaffected in such circumstances. The resilience of this buffering also becomes seriously challenged as the duration of food insecurity gets prolonged. This is especially noticeable in the low income urban households. In our study, chronic food insecurity increased significantly among urban adolescents who were members of food insecure households as income declined from high to low tertiles. Such a trend was not observed among rural and semi-urban adolescents. This finding suggests that the protection of adolescents from chronic food insecurity diminishes especially in the urban households as purchasing power gets further constrained.

As adolescents are in a phase of rapid growth and development the impact of chronic food insecurity on their well-being is considerable. In food insecure situations, as food becomes the first priority, expenditures on other goods and services could also be forgone to spare money for buying food [12]. This exposes population groups with the high nutritional requirements, including young children, adolescents, pregnant and lactating women to malnutrition [10]. Reductions in dietary quantity and quality are associated with many adverse health consequences [19, 34–39], and these impacts may be magnified when food insecurity is chronic.

The study also demonstrated that a larger proportion of girls were chronically food insecure than boys in all residence places. This findings supports a male gender biased buffering of adolescent against food insecurity reported from the same cohort [16]. In developing countries boys are often favored in the allocation of resources within the household for cultural reasons [40–44].

Higher educational status of adolescents was negatively associated with the likelihood of having chronic food insecurity, only for urban adolescents. Education contributes to both income and knowledge and those adolescents who had better education may have better access to employment and hence to food. The strong relationship between food insecurity and lack of education has also been well documented by other studies [45, 46]. However, in this study although paternal and maternal educations were negatively associated with chronic adolescent food insecurity, the association disappeared in multivariable analysis, suggesting that its effect is mediated through other variables including household income.

Our analysis also showed that a high dependency ratio was an independent predictor of chronic food insecurity among adolescents across all urban, semi-urban and rural residences. When the number of dependents is high compared to income earners, household level food availability is constrained due to increased consumption compared to availability of food. Coupled with an increase in the cost of food items, the purchasing power of the family will not be sufficient to obtain enough food for all members. Food assistance programs could play a protective role for children and adolescents in low income households in such circumstances [47]. A report from Southern African case studies by Devereux 2002 [48] showed that “even tiny income transfers invested in income-generating activities, education, social networks, or the acquisition of productive assets can make a difference”, suggesting that social safety nets can play a significant role in reducing poverty and chronic food insecurity. Food insecurity is a complex phenomenon and varies through a continuum of successive stages as conditions become more severe. Therefore, there is a need for timely intervention before the situation further deteriorates.

Seasonal variation of food availability could affect the level of food insecurity in the study population. However, as the data were collected over three seasons, and each adolescent was interviewed about food security levels at the same season of the year, seasonality should not have a noteworthy effect on the findings. Our analyses also showed that there was no significant difference in chronic food insecurity between seasons (Table 1).

This study used longitudinal data to document predictors of chronic food insecurity in the context of rising food prices in Ethiopia and highlighted how increasing food prices erode intra-household buffering (protection mechanisms). However, although the food security items were adopted from scales used in developing countries and have high internal consistency [Cronbach’s alpha = 0.81], the fact that they are not validated locally might lead to a problem of misclassification. However, the fact that they vary with socioeconomic variables consistently in a plausible way indicates that this possibility is remote.

The results showed that adolescents from urban areas who are members of low income households are at greater risk of chronic food insecurity. An increase in the proportion of chronically food insecure adolescents only in urban areas as income declines from high to low tertiles suggests that the resilience of buffering is eroded as purchasing power diminishes. Female sex, primary level education, high household dependency ratio were also independent predictors of chronic adolescent food insecurity regardless of place of residence. Food security interventions including safety net and market stabilization programs should target the urban low income households to reduce the extent of chronic food insecurity among adolescents.