COMCEC

Malnutrition in the OIC Member

Countries: A Trap for Poverty

potential and at risk of becoming short adults (e.g. WHO, 1992). Not all the intergenerational

transmission of malnutrition is channelled through the effect of low birth weight. Provision of

nutritious food and control of infections (Golden 1998, Martorell et al. 1994), as well as

improvements in socio-economic status across generations (Hauspie et al, 1996) would allow

the child to partially catch-up the effect of LBW. In contrast, insofar as the short stature of

mothers stem from lack of access to healthcare or nutritious food, or the implementation of

harmful feeding practices, then the causes behind maternal malnutrition will also exert their

impact on child growth, beyond the independent effect of LBW.

Remarkably, very little differences were found when both measures of maternal malnutrition

are introduced together in the regressions, which suggests that LBW is far from channelling all

causes of intergenerational of malnutrition. The regression results when both variables are

included together were thus be the ones reported.

Logit regressions and odds ratios

Each of the three malnutrition variables that will be used as dependent variables in the

regressions are of a binary nature (they take the values 0 or 1). To estimate this type of binary

variables, the logit estimator is used. The logit estimator constraints the predicted value of the

dependent variable to be comprised between

and

Logit regressions yield coefficients that are non-linear with the value of other regressors and

which are hard to interpret. The estimated coefficients will therefore be presented in the odds

ratios form (which corresponds to the exponentiated value of the coefficients). Odds ratios are

constant with the value of all covariates and have a simple interpretation: they indicate by how

much the odds of the dependent variable to take the value

change when the covariates

increase by one unit (or switch to positive status if the covariate is a binary variable). An odd

ratio of

indicates that the likelihood of malnutrition does not change with the covariate, i.e.

that there is no relationship between the two variables. An odd ratio greater (lower) than

indicates that the covariate is positively (negatively) associated with malnutrition.

Stakeholders' interviews and review of secondary literature

To complement the quantitative analysis, a number of stakeholders' interviews have been

carried out in each country (usually around 8-10). These interviews followed a semi

structured pattern, with the core of the interview focusing on the areas of intervention and

expertise of the respondent. The same interview protocols were used in each country,

although some questions were tweaked to better reflect local realities. The interviews were

especially meant to shed light on the challenges and successes of policies aimed at addressing

malnutrition. The existing literature (reports, books and peer-reviewed articles) was also

reviewed to provide further context and insights to the analysis.

Integration of quantitative and qualitative research

Each case study follows the same template. They start by presenting data on levels and trends

of malnutrition before turning to the analysis of the links between malnutrition and poverty.

To that end, results of the quantitative analysis are shown, and discussed in the light of key

findings of the review of secondary literature and the stakeholders' interviews. To achieve a

consistent structure throughout, the analysis of poverty and malnutrition is always broken

down into (i) the overall role of poverty, (ii) food security, (iii) health, water and sanitation and

(iv) IYCF practices and breastfeeding. These do not exhaust all the potential pathways through

which poverty and malnutrition are related but they proved extremely relevant in all case