Older people, nutrition and emergencies in Ethiopia
By Vanessa Tilstone
Vanessa Tilstone has worked for HelpAge International in Ethiopia for the last 3 years as the Country Programme Director and has worked previously in Mozambique, Ethiopia, Malawi and Brazil. Dolline Busolo is the Regional Nutritionist for Helpage and has been based in Nairobi for the last 2 years.
This article highlights HelpAge International's work in researching and promoting the needs of older people in emergencies and in developing tools for nutritional assessment of older people1. Using the example of its Ethiopia country development programme it focuses on the key debates and challenges and makes recommendations for future action.
The situation of older people in Ethiopia
Old woman in Borena, Ethiopia (2000)
There is an underlying respect for older people in Ethiopia where family and community support systems are relatively strong. However, a significant number of older people have no family and community support, mainly due to the death of relatives or separation caused by famine, war, disease and displacement and the weakening of family and community support structures.
Even in family settings, older persons are often disproportionately affected in times of drought. Coping strategies are often limited as they may not be able to travel long distances in search of pasture, water or food or engage in daily labour or other income generating activities. They may have difficulty consuming wild or drought foods. In addition, they may also sacrifice themselves in order to save the lives of other members of the family, either by refusing food, eating last or preferring to be left behind when families migrate (see Case Study 1).
These situations are usually not identified by aid agencies. Vulnerable older people, particularly the bedridden or itinerant beggars, are not easily visible and are rarely prioritised for interventions. Until recently, nutritional surveys have focused exclusively on children under five (see Case Studies 2 & 3).
The Ethiopian government's policy on food aid targeting excludes people thought to be over 60 years from participating in Employment Generation Schemes (food for work), regardless of their physical ability or 'real' age. General food distributions (GFD), if carried out at all, do target vulnerable groups including older persons but only account for 20% of food distributed. In practice older people, particularly women, are often excluded. They are often not regarded as permanent residents in the areas where they presently live as they have moved from their homes due to widowhood, poverty or crises. Supplementary feeding focuses on children under five years and lactating mothers. Older people are not included in this intervention despite the fact that this group often have poorer nutritional status and great difficulty digesting the coarse wheat and lentils that are usually distributed in the GFD.
Assessing older people's nutritional status
A number of difficulties were experienced promoting the inclusion of older people in emergency programmes, including a belief that older people are an unproductive group and will always be cared for by relatives, regardless of the food stress. By far the biggest challenge, however, was the need of further research on assessment of malnutrition, among adults and older people.
Initial research by HelpAge International recommended that armspan be used as a proxy for height in Body Mass Index (BMI) measurements for older people. However when the armspan equations were tested in Ethiopia it was found that there were differences from other African populations. A research project was conducted among the three main ethnic groups and a fourth group which was particularly affected by drought. The regression equations calculated were found to be different for each of them.
Overall, a number of difficulties were experienced using BMI as an indicator of nutritional status (see Table 1). By far the biggest problem was the need to adjust for Cormic index (Norgan's correction) which takes into consideration the fact that different phenotypes have different sitting:standing height ratios which will affect their BMIs. This is a complex equation that needs to be derived for each ethnic group.
|Table 1 Problems using BMI2.
|Measurement difficulties and risk of error
- Mathematics required for BMI and regression equation
- Error in height squared
- Confounded by famine oedema
- Height difficult to measure, especially in older people
- Height derived from a proxy (demi-span/arm-span), error
|Populations have different phenotypes (body-shapes)
Interpretation affected by phenotype
Need to determine population specific Cormic Index and adjust using Norgan's correction.
Age related changes (BMI decreases with age, fat redistribution, height decreases with age)3
|Lack of baseline information
||Normal prevalence rates, seasonal patterns and trends unknown
|Need additional information for interpretation of findings
||Findings cannot be interpreted in isolation of other information
Note: BMI is calculated as Weight [kg] ÷ (height [metres])2
The Mid Upper Arm Circumference (MUAC) was found to be a simpler and quicker measurement. MUAC cut-off points had been determined in the initial research by London School of Hygiene and Tropical Medicine (LSHTM), i.e. <22cm for severe and <23cm for moderate malnutrition. However, these are for populations in stable nutritional situations and were not found useful for determining the need for intervention in emergency situations. Collins et al. had determined lower MUAC cut-off points for screening for nutritional interventions in South Sudan, where, among the Nuer people, over 94% were considered malnourished if BMI measurements alone were used.
In Ethiopia these cut-off values were found to be very low and therefore a middle ground, based on Action Contre Faim's (ACF) experience of therapeutic feeding in Rwanda, was used. This is being used in two interventions currently being supported by HelpAge International Ethiopia in Werder and Somali region. However, as the case studies show, there is still much debate about which cut-off values should be used. A summary of the three recommendations is outlined in Table 2.
|Table 2 MUAC cut-offs for adults
|A. Characteristics of the 85 clients alive at 12 months
||Ismail et al. (HAI/LSHTM)3
||Collins et al. (July, 2000)4
* with or without clinical/social criteria or
** with social/clinical criteria
There is an increasing acceptance that older people are particularly vulnerable in many areas of Ethiopia in times of drought and conflict. A total of ten NGOs have included older people in their nutrition surveys and at least five interventions have been carried out to address older people's needs. For the first time in 2001 the national government food appeal mentioned older people as a priority for supplementary food and the NGO umbrella organisation, CRDA (Christian Relief and Development Association), has included increasing advocacy for older people as one of the objectives of its emergency task force.
However, more research is required on the vulnerability of older people and nutritional assessment including:
- Whether BMI measurements need to be adjusted for sitting height (Cormic index) and its calculation for various ethnic groups;
- Refinement of MUAC measurements for adults and older people for severe and moderate malnutrition by relating MUAC to functionality and morbidity;
- More research on how anthropometric measurements vary with age;
- Ethnic specific armspan to height ratios to use armspan as a proxy for height for older people in BMI measurements;
- More research into the vulnerability of older people and access to relief interventions in different regions and contexts;
- Increased awareness within aid agencies of older people's vulnerability and their right to life.
Nutrition guidelines for older people need to be incorporated into agency guidelines. From experience so far, the main lessons learned are:
- Qualitative information on older people, such as underlying causes and risks of malnutrition, and changes in support structures, should be systematically collected during nutritional assessments. Further data on the effects of food aid use and intra-household food sharing patterns on older people are required.
- For rapid assessments where nutritional status cannot be assessed using sound sampling procedure, it is more appropriate to collect good qualitative information rather than taking MUAC measurements of a small convenient sample.
- The index and the cut-off values used (and the correction process if applicable) should always be defined when reporting prevalence rates.
- Prevalence rates should always be interpreted in the context of other information e.g. food security, malnutrition rates among children under five, information on social and community support structures.
- Older people need to have access to foods that are easily digestible and provide adequate amounts of micronutrients. Improved access to blended foods may fulfil these requirements.
It is hoped that agencies working in Ethiopia and other countries facing humanitarian emergencies and natural disasters can take up these issues so that older people's needs are addressed more effectively.
Case Study 1
Coping strategies that discriminate against older people, Yabello, Borena Ethiopia 2000- 2001 (Vanessa Tilstone, HAI)
In Borena, Southern Ethiopia, during food shortage periods, older people voluntarily refuse food in order that other family members can survive. This practice is in addition to the normal preferential feeding of children also characteristic in this culture. In 2000, Norwegian Church Aid reported that older people were refusing food, only consuming liquids and conserving energy by resting for long hours. A nutritional survey carried out by HAI in April 2000 estimated a prevalence of 54% global malnutrition among older people (using unadjusted BMI as an indicators (BMI less than 17)). In June 2000, GOAL included older people in their survey, showing 64% of those assessed with MUAC values less than 23cm using MUAC.
Despite their widely recognised vulnerability, older people were not prioritised in interventions until May 2001, when HAI supported GOAL to implement a supplementary feeding programme focusing on older people. In July 2001, 510 older people, 164 children, and 36 pregnant and lactating mothers were registered for supplementary food from Yabello and Teltelle Woredas when they were found to be moderately or severely malnourished. Criteria for admission were:
- Reported age over 55 years; or nearest historical event.
- Physical signs such as gray hair, wrinkled skin etc.
- Cross checking with elders,
- MUAC less than 21cm.
Case Study 2
Including older people and adults in nutritional surveys: Damot-Wede, Ethiopia, 2000 (Kate Sadler, Concern Worldwide)
Concern has been implementing emergency nutrition programmes in Damot Weyde since April 2000, when a first survey showed a global malnutrition rate of 25.7% among children under five years old. As a result supplementary and therapeutic programmes were set up. Reports of sick adults presenting for treatment led Concern to include adults and older people in their nutrition survey in July 2000 in order to estimate the extent of the problem.
Borena, Ethiopia (2000)
Data collected included sex, age, oedema, weight, height, sitting-height and MUAC. A measurement for sitting height was taken to determine the Cormic Index (body shape) for the population. The Cormic Index was calculated from the ratio of the sitting height to standing height. For adults whose height could not be measured, no data were collected. The analysis used two age categories; 18-49 and >49 years. The prevalence of malnutrition was reported for observed BMI and adjusted BMI. The adjusted BMI took into consideration the Cormic Index for the population being surveyed.
The results showed large differences between the prevalence rates reported for observed BMI and adjusted BMI. Using a BMI cut-off of <17kg/m2, the prevalence of malnutrition among younger adults (18- 49 years) was 1.7% (adjusted for Cormic Index) and 10.7% (unadjusted). The prevalence of malnutrition among older people (>49 years) was 2.0% and 24.5% for adjusted and unadjusted prevalence rates respectively. The large differences in the prevalence rates can be attributed to the fact that the study population has a relatively low Cormic index (long legs for body stature in relation to trunk length). Therefore without applying the Cormic Index adjustment factor, the BMI is artificially lowered and results in a much higher proportion of the population being reported as malnourished. The Cormic Index for this population was calculated to be 0.5 compared to 0.52 for the reference population.
The survey determined that a MUAC cut-off of 18cm identified a similar proportion of adults with global malnutrition to that using an adjusted BMI of <16 (severe malnutrition). But further tests are required to see if the same population groups were affected. The survey also showed that mean BMI and mean MUAC values significantly decreased with increasing age.
There were a number of additional constraints that the survey encountered. These included: (1) sitting height needed to be monitored closely (very easy to be inaccurate) and the additional measurement slowed the survey team down considerably; (2) the adjustment process required specific expertise and (3) the sample size of older people was too low.
Case Study 3
Including older people in a Supplementary Feeding Programme in Bolosso Sore 2000 (Laura Phelps, Oxfam-GB)
Many of the older people in Bolosso Sore are marginalised socially. Chronic health and social problems exacerbate their poor nutritional status. The government DPPC (Disaster Prevention and Preparedness Commission) beneficiary selection criteria for gratuitous and Employment Generation Schemes does not allow for displaced people as beneficiaries have to be 'permanent residents'. As a consequence, Oxfam included older people in their Supplementary Feeding Programme (SFP).
In November 2000, there were 208 older people (more than 50 years old) registered in the SFP, of which 98% were female. At least 95% of the target population were living in the urban centres of the woreda. The criteria used for selection was MUAC <18.5 cm. Assessment using BMI showed that none of the admissions had a BMI <17.0.
An assessment of the causes of malnutrition for older people in Bolosso Sore indicates that this proportion of the population is chronically vulnerable. Many of the older beneficiaries in the SFP were widows and the majority had no access to land. Poor use of food is exacerbated by physiological problems, especially sight and dentition difficulties as well as chronic illness. Community support was strong where families and relative are close by. Once the older people are separated and have to move to new areas, community support was notably poorer. All the women were economically unproductive and received small income from begging outside the church or market.
View the postscript that this article relates to
1Older people in disasters and humanitarian crises: Guidelines for best practice (HelpAge International, London, 2000); Suraiya Ismail & Mary Manandhar, Better Nutrition for Older People: Assessment and Action (HelpAge International/London School of Hygiene and Tropical Medicine, 1999).
2A. Borrel, Report of workshop on addressing the nutritional needs of older people in emergencies, November 2000.
3Ismail & Manandar op. cit.
4Collins, S., Duffield, A. & Myatt, M. Anthropometric Assessment of Nutritional Status of Adults in Emergency- Affected Populations July 2000 Geneva ACC/SCN.
5Yvonne Grellety, Personal communication.
Taken from Field Exchange Issue 14, November 2001