milk matters: a literature review of pastoralist nutrition and programming responses

8/14/2019 Milk Matters: A Literature Review of Pastoralist Nutrition and Programming Responses

1/50

A Literature Review of Pastoralist Nutritionand Programming Responses

Kate Sadler, Carol Kerven, Muriel Calo,Michael Manske and Andy Catley

February 2009

MILK MATTERS


2/50

Suggested citation

Sadler, K., Kerven, C., Calo, M., Manske, M. and Catley, A. (2009). Milk Matters: A literaturereview of pastoralist nutrition and programming responses. Feinstein International Center, TuftsUniversity and Save the Children, Addis Ababa.

Acknowledgements

This review was made possible with funding from the Office of Foreign Disaster Assistance (OFDA)United States Agency for International Development (USAID). The authors would like to thank TinaLloren, Alemtsehay Greiling and Adrian Cullis of Save the Children USA and Matthew Hobson,Themba Nduna and Abdirahman Ali of Save the Children UK for providing considerable feedbackand support for this review. Thanks also go to Fiona Flintan and Yacob Aklilu for provision of referenced literature, to Anastasia Marshak for help with graphics and to Alison Napier and JonelleLonergan for proof reading and edits.

Photo credits

Dawit Abebe p34; Yacob Aklilu p35; Suzan Bishop p29; Andy Catley p10, 19;Fiona Flintan cover photo; Cara Kerven p8, 11(right), 14; Kelley Lynch p33; Michael Wadleigh p11(left).

Cover page graphicsChild drinking from gourd reproduced from training materials developed by VSF Belgium in SouthSudan; mud cow in South Sudan adapted from line drawing by Evans-Pritchard, 1940.


3/50

C ONTENTS

1. I NTRODUCTION 1Background to Save the Childrens Africa Region Pastoral Initiative 1 Aim of Milk Matters and the literature review 1

2. M ETHODS 1

Literature search 1Focus of the literature review 23. E PIDEMIOLOGY OF M ALNUTRITION IN P ASTORALIST S ETTINGS 3

A note on child growth 3Seasonal fluctuations 4Summary - epidemiology of malnutrition in pastoralist settings 5

4. M ILK AS A F OOD FOR C HILDREN 6Summary milk as a food for children 9

5. P ASTORALIST R ELIANCE ON L IVESTOCK M ILK FOR N UTRITION AND L IVELIHOODS 9 Why pastoralists rely on milk 9Livestock herding strategy 10The contribution of milk and milk products to energy and nutrient requirements 11Hard choices: milk for calves, milk to eat, milk to sell or milk for social exchange 12

- Impact of commercialization on the use of milk 13The role of milk in the diets of infants and young children in pastoralist settings 14Summary - pastoralist reliance on livestock milk for nutrition and livelihoods 15

6. H OW P ASTORALIST D IETS AND L IVELIHOODS A FFECT N UTRITIONAL 15 S TATUS Diets and nutrient intakes 15

- Reliance on milk: a protective or harmful factor? 15- The effect of season on nutrient intake and nutritional status 17- The effect of animal disease on milk availability and nutritional status 18

Understanding infant and young child feeding practice and its consequences 20The links between maternal and child nutritional status 21Sedentarization and commercialization 22 Wealth 23Morbidity and health care coverage 24Summary - how pastoralist diets and livelihoods affect nutritional status 25

7. P ROGRAM STRATEGIES USED TO ENHANCE MILK SUPPLY AND CONSUMPTION 26Breast milk and animal milk: Improving feeding practice for young pastoralist 26childrenLivestock development and humanitarian interventions 26

- Improving animal health to increase milk supply to children 27- Increasing livestock ownership to provide milk to children 30- Water, feed, pasture and range management 32

Traditional milk preservation 34Milk and markets 34Other strategies 35

- Commercial destocking as a response to drought 35- Cash distributions during drought 35- Food aid distributions 35- Safety nets 36

Summary - program strategies used to enhance milk supply and consumption 368. C ONCLUSIONS 36References 38


4/50

Acronyms

ARI Acute Respiratory Infection ASF Animal Source FoodsCAHW Community Animal Health WorkerCHW Community Health Worker

EAR Estimated Average RequirementENN Emergency Nutrition NetworkFAO Food and Agricultural OrganizationFGD Focus Group DiscussionFIC Feinstein International Center, Tufts UniversityFMD Foot and mouth diseaseFSAU Food Security Analysis Unit, SomaliaIDP Internally Displaced PersonILRI International Livestock Research InstituteMUAC Mid-Upper Arm CircumferenceNGO Non-Governmental OrganizationODI Overseas Development InstitutePHNI Pastoralist Health and Nutrition InitiativeRNI Recommended Nutrient IntakeSCUK Save the Children UK SCUS Save the Children USASD Standard deviationSSR Standing height to sitting height ratioWFH Weight for HeightWHO World Health Organization


5/50

1

1. INTRODUCTION

Background to Save the Childrens AfricaRegion Pastoral Initiative

Created by Save the Children USA in mid-

2007, the African Region Pastoral Initiativeaims to develop the evidence base for programming in pastoralist settings; to usethis experience to advocate for better practices,

programs and policies; to support effectiveimplementation; and to work with the SC

Alliance, communities, government and otherstakeholders to take the work to scale .1

One component of the Regional Initiative isthe Pastoral Health and Nutrition Initiative(PHNI), created in April 2008 to helpunderstand and prioritize interventions thatimprove the health and nutritional status of children in pastoral settings. Within thePHNI, Save the Children USA and UK (SCUS/SCUK) and the FeinsteinInternational Center (FIC) at Tufts Universityhave joined efforts on a venture called MilkMatters to explore interventions related to oneof the most important component of childrensdiets: human and animal milk.

Aim of Milk Matters and the literature review

In the first phase Milk Matters is primarily anEthiopian initiative, with the agenciesinvolved all based in Addis Ababa. Therefore,the activities under the following aims arefirst and foremost focused on the Ethiopianpastoralist population, but will, we hope,create learning to be applied to the broadercontext of pastoralists in the African region.

To identify interventions that lead toan enhanced supply of animal andbreast milk for children in pastoralcommunities across seasons andwealth groups, thereby leading to

improved nutritional status; To support implementation of effectiveinterventions at a large-scale;

To advocate for policies that promotehealthy and well-nourished children inpastoral communities and influencethe implementation of the

1 From Terms of Reference for Literature Review July2008.

Governments national NutritionStrategy accordingly;

To communicate the results to a broadaudience of pastoral communities,government institutions, researchersand practitioners.

A great deal is already known about thenutrition, health and food security of childrenin pastoralist populations. The aim of thisliterature review was to draw together thisknowledge, particularly that around the roleof milk in the livelihoods of pastoralist people;and to help us understand those aspects of pastoral child nutrition that are wellestablished and those issues that remaindebated or poorly covered by the literature.

Broad themes to be investigated were: The epidemiology and causes of

malnutrition in children in pastoralistcommunities; including debate on howwe measure malnutrition in thesecommunities;

The role of human milk (i.e.breastfeeding) and of animal milks inthe diet of children in pastoralistcommunities;

Key interventions that have aimed toimprove breastfeeding andcomplementary feeding practice inpastoralist regions and their impacton nutritional status of children (if this is measured), and;

Key interventions that have aimed toenhance milk supply from animals inpastoralist regions and their impacton nutritional status of children (if this is measured).

2. M ETHODS

Literature search

Major reference databases weresystematically searched using a set of agreedsearch terms and key words. These includedISI Knowledge, Expanded Academic ASAP(Gale), Medline (OVID), CAB Abstracts, andPubMed. Google Scholar and the onlinepublications databases for the ENN FieldExchange and the ODI Humanitarian PracticeNetwork and Pastoral Development Network


6/50

2

were also searched. Articles andbibliographies received from FIC and SCUK were reviewed for relevance to the literaturereview topics. These sources were found toinclude many NGO draft materials,evaluation documents and field reports

related to the objectives.References were organized by theme and sub-theme linked to one or more of the aboveobjectives and labeled accordingly under thekeyword heading in the reference database forready sorting. Articles available electronically(the vast majority) were downloaded andcategorized into folders labeled by theme. Afew hard or scanned copies were obtainedwhere electronic versions were not available,either from the library stacks or fromneighboring university libraries, usinginterlibrary loan services. Citations andabstracts for each reference were uploaded ortyped into reference management software(Endnote 2006).

Themes originally developed by the literaturereview team and included in the first draft of reports were subsequently consolidated intothe themes presented in this report.

Focus of the literature review

Pastoralism is defined in many ways. In 1994Fratkin and Smith defined it as:

a strategy in which people raiseherd animals, often in arid ormarginally agricultural regions, inorder to provide a regular supply of

food in the form of milk, meat, bloodand trade for grains. For the most

part, this production is subsistencebased, aimed at producing foods(primarily milk) for householdmembers. Exchange is secondary

(Fratkin and Smith 1994). Although this definition conforms to theclassically held view of pastoralism, manywould agree that it is now rather outdated,and that exchange plays a key role in manypastoralist livelihood strategies (see sections 5and 6). A more current definition has been putforward by Nori and Davis:

Pastoralism is the finely-honedsymbiotic relationship between localecology, domesticated livestock and

people in resource-scarce, climaticallymarginal and highly variableconditions. It represents a complex

form of natural resource management,involving a continuous ecologicalbalance between pastures, livestockand people (Nori and Davies 2007).

The primary focus of this review is pastoralistethnic groups whose livelihoods fall under thedefinition above. However, in many casesthere are rapid socioeconomic changeshappening within these groups that arechanging livelihood strategies and reducingreliance on animals for food. Consequently,some of the literature refers to communitiesthat might now be better described as agro-pastoralist i.e. with livelihoods that depend ona mix of agriculture and raising animals.However, there are common elements withinboth pastoralist and agro-pastoralist groups;including an element of mobility, a livelihoodstrategy that maximizes milk and meatproduction and a diet that includes relativelylarge amounts of animal products; that arelikely to impact the nutritional status of children. Most of the findings of this reviewtherefore, could be applied equally to bothlivelihood groups.

Pastoralist communities can be found all overthe world, from the mountains of the HinduKush to the deserts of Sudan. On the topic of milk, the Maasai of Kenya and Tanzania areby far the most well-researched pastoralistpeoples globally; a total of 10 publicly-available series of studies exist. Kenya is alsoa common research location, with 20 (nearlyone third) out of 67 country-specific studies.

Yet by the end of the 20 th century, milk wasbeginning to play a smaller role in the diets of

some of these ethnic groups, such as theMaasai of Kenya. Several other East Africanpastoral groups have also been extensivelyresearched regarding milk and nutrition; forexample, the Turkana of northern Kenya,stemming from the STEP long-term researchprogramme initiated by Jim Ellis, andsecondly the Boran of southern Ethiopia withnumerous studies by staff from ILCA, now


7/50

3

ILRI (the International Livestock ResearchInstitute). In the rest of the world, there havebeen very few studies of milk from the herdsand flocks belonging to the millions of pastoralists still extant in Asia, the MiddleEast and South America. This literature

search located only 10 studies with anyinformation on livestock milk in pastoralistsocieties outside Africa.

The report presented here focuses primarilyon pastoralist ethnic groups in eastern Africa,as it is this literature that is most useful forthe broader objectives of the PHNI. However,the literature from other regions of the worldmay in some instances provide usefulcomparisons and contrasts to the role of livestock milk in East African pastoralistsystems. Therefore, it may be advantageous torevisit this literature at some point during theproject.

Lastly, it is important to note that there existshuge variation in how different pastoralistgroups manage their animals, secure food andfeed their children. This review attempts todraw together the most important works thatgo some way in explaining these aspects of pastoral livelihoods, but it will be importantfor the PHNI to do further work to increaseunderstanding of particular issues withinthose ethnic groups targeted by the initiative.The primary data collection planned as afollow up to this literature review should helpto do this.

3. E PIDEMIOLOGY OF MALNUTRITIONIN P ASTORALIST S ETTINGS

Nutrition studies and surveys over the pasttwo decades have shown that children inpastoralist populations present with higherlevels of wasting (measured by weight forheight [WFH]) and lower levels of stunting(measured by height for age) than agrarianpopulations in the same countries and regions(Wagenaar-Brouwer 1985; Little, Gray et al.1993; Sellen 1999; Roth, Fratkin et al. 2003;Fratkin, Roth et al. 2004; Gray, Wiebusch etal. 2004; Fratkin, Nathan et al. 2006; Myatt2007). Pastoral child populations in theGreater Horn countries of Eritrea, Ethiopia,Kenya, Somalia, South Sudan, and Uganda

demonstrate wasting prevalence of 17% onaverage, seven percentage points higher thanagriculturalists or populations with mixedlivelihoods in the same regions. (Mason,Chotard et al. 2008).

A note on child growth

Recent work has suggested that some of thisdifference in levels of wasting in children agedbetween 2-5 years may be due to differences ingrowth or body shape rather than differencesin actual nutritional status (Myatt 2007).Children in pastoralist communities had, onaverage, a significantly lower Cormic Index(or sitting height to standing height ratio[SSR])2 than children in agrarian populations.This had the effect of overestimating theprevalence of wasting measured by weight forheight in pastoralist children in this agegroup. This effect is a well establishedproblem with using the body mass index toassess the nutritional status of adults (Collins,Duffield et al. 2000). However, this finding forchildren under five years conflicts with that of the World Health Organization, which, indeveloping its new growth referencepopulation and standards, maintains that allchildren under the age of five demonstratesimilar growth (weight to height ratio) (deOnis, Garza et al. 2003). Myatts work showedthat levels of wasting measured by mid-upperarm circumference (MUAC) on the other handwere not associated with SSR (Myatt 2007).

The possible reasons behind a differentpattern of growth among children under fiveyears are debated. The most common school of thought relates it to diet, with several authorsdiscussing the relationship of growth patternto a diet high in protein (particularly milk)and other nutrients that support growth instature (see section 4), but low in energy(particularly during leaner times) that does

not support deposition of soft tissue (or fat).This results in pastoralists demonstratinggreater fluctuations in wasting (particularlyin lean years), as well as higher baselinelevels (Mason, Chotard et al. 2008). Otherstudies suggest that linear morphology has a

2 SSR - Sitting Height/Standing Height. Smaller values of SSR indicate longer limbs and/or shorter trunks. Largervalues of SSR indicate longer trunks and/or shorter limbs.


8/50

4

strong genetic component arising out of selection pressures related to diet. Metabolicadaptations to the pastoralist diet are thoughtto occur during the growth period and inparticular during the weaning transition,when butterfat given as a complement to

breast milk is replaced by protein as theprincipal source of energy. Selection formetabolic plasticity is intense at this time,and the assumption is that there is acharacteristic pattern of growth in survivingchildren, in which growth in height andweight become increasingly discordant. Thiscould mark the attainment of metaboliccompetence among these individuals (Gray1996; Gray 1998; Gray, Wiebusch et al. 2004).

Authors differ in their suggestions on how todeal with this. Myatt (2007) recommendsmoving to a different indicator to measurenutritional status in pastoralist populations.He refers to previous work that hasdemonstrated the relationship betweenMUAC thresholds and short-term mortalityrisk - a relationship that is found not to varysignificantly between populations. For thisreason the author concludes:

In the populations represented in thisstudy, MUAC is likely to be the mostrobust (i.e. to body-shape) indicator of acute under-nutrition. If WFH is usedthen, as with adults, it should becorrected for body-shape beforeapplying case-definitions. Thecorrection for body-shape would,however, require the collection of sitting height which would complicatesurveys whereas the use of MUAC would considerably simplify surveys.

Other work (Grobler-Tanner 2006; Mason,Chotard et al. 2008) has suggested that,because of the higher prevalence of wasting as

measured by WFH in pastoralist populations,we continue to use WFH but raise levels of wasting that:

a. Define the baseline (or the norm)from 5-10% in agrarian/mixedpopulations to 15-17% in pastoralistpopulations;

b. Trigger emergency food interventionsfrom 15% in agrarian/mixed

populations to 20-25% in pastoralistpopulations.

Neither of these recommendations howeverappears to be based on any functionalmeasure of outcome related to nutritional

status i.e. how does a change in the way wedefine baseline levels of malnutrition relate tooutcomes such as mortality and morbidity.

Seasonal fluctuations

Seasonality is a crucial factor in the study of pastoralism with livelihood strategiesrevolving around seasonal patterns of waterand plant availability and correspondinggeographic mobility. There is without doubt aseasonal increase in vulnerability tomalnutrition when livestock pasture becomesscarce, in turn limiting both drinking waterand milk availability for human consumption(Fujita, Roth et al. 2004). However, seasonalfluctuations in nutritional status are noted tovary considerably by pastoralist ethnic group.Studies among the Turkana and Datogadocument only a moderate decline inanthropometric status in the dry season,despite significant seasonal variation inrainfall and food availability (Shell-Duncan1995; Sellen 2000). The Rendille and Ariaalstudies by Fratkin, Nathan and Roth describeless variability during the dry season amongpastoralists compared with their settledcounterparts, and suggest that in thesecommunities, seasonally determined foodavailability is not a significant determinant of nutritional status (Fratkin, Roth et al. 1999;Fujita, Roth et al. 2004). A more significantdecline in nutritional status by season hasbeen noted among the Fulani and Tamasheqin Niger (Loutan and Lamotte 1984;Wagenaar-Brouwer 1985), the South Turkanaethnic groups (Galvin 1985; Little 1989) andthe Somali pastoralists in Ethiopia (SCUK

2007).In Eastern Africa this decline is not usuallylinked to the cereal hunger gap, but occurs inthe early half of the year linked to the dryseason, the end of which is considered to bethe most dangerous period for human healthin pastoralist societies (Fratkin, Roth et al.1999) due to decreased pasture, reduction in


9/50

5

milk production and increased work loads.However, the size of the seasonal effects, of

about five percentage points in normal years,are not such as to obscure the considerablylarger shifts seen in years of deterioratingfood security i.e. in this region also thenutritional state of pastoralist groups seem tobe relatively protected from seasonal changesin food availability (Mason, Chotard et al.2008). This should not obscure the fact thatother responses to seasonal reductions in foodavailability, the slaughter or sale of livestockfor example, may be increasing vulnerabilityin other ways.

Summary - epidemiology of malnutrition inpastoralist settings

Pastoralists demonstrate higher levelsof wasting measured by WFH atbaseline than agrarian communities.Recent work suggests that this, atleast in part, is due to different growthpatterns of pastoralist and agrarianethnic groups and that MUAC shouldbe used in preference to WFH toassess nutritional status of pastoralists. Further study is requiredthat links levels of malnutrition(measured by weight for height and by

Table 1: Nutritional status in Somali Region, 2006 (adapted from SC UK, 2007)

Date Season Pastoral area Prevalence of globalacute malnutrition(95% confidence interval)

Prevalence of severeacute malnutrition(95% confidence interval)

January, 2006 Dry Dollo-ado, Dollo-bay, Barrey 18.8% (15.8%, 21.6%) 1.4% (0.3%, 2.5%)

August toSeptember, 2006

End wet/early dry

Dollo-ado, Dollo-bay, Barrey

14.5% (11.8%, 17.2%) 0.7% (0.1%, 1.3%)

January, 2006 Dry Moyale, Hudet 19.7% (16.0%, 23.4%) 1.8% (0.8%, 2.7%)

August toSeptember, 2006

End wet/early dry

Moyale, Hudet 7.6% (5.3%, 9.8%) 0.1% (0.0%, 0.3%)

Figure 1: Seasonal fluctuations in wasting measured by WFH in Somali Region, Ethiopia(adapted from Mason, Chotard et al. 2008)

0

5

10

15

20

25

30

1 2 3 1 2 1 2 3 1 2 3 1 2 3 1 3 3

Seasons (2000-2006): 1 - moderate; 2 - hunger; 3 - post harvest.

P r e v a

l e n c e o f w a s t

i n g

( % )


10/50

6

MUAC) with functional outcomes suchas morbidity and mortality in pastoralpopulations. This will help defineappropriate trigger levels forhumanitarian intervention.

The hungry or lean season in

pastoralist populations is oftendifferent from that of agrariancommunities and appears to be linkedto the dry season and a reduction inavailability of pasture andsubsequently milk, rather than acereal gap.

In some ethnic groups, the nutritionalstatus of pastoralist populationsappears to be relatively protected fromseasonal changes in food availabilityin normal years, but appears tofluctuate more dramatically in yearsof drought and crisis. This hasimplications for the timing of nutritionand health intervention and the typeof intervention in normal and crisisyears.

4. M ILK AS A FOOD FOR CHILDREN

The importance of milk for infant and childhealth is well established. In Europe, thenutritional value of milk was repeatedlydemonstrated by intervention studies between

the two world wars and, in the face of considerable wartime undernutrition, the firstNutrition Policy for Britain (1939-1945) reliedheavily on milk to provide the extra nutrientsto those with special physiologicalrequirements; mainly young children andpregnant and lactating women (Darke 1979).The general consensus was that thisnutritional policy was highly successful andcontributed to a considerable improvement inthe health of the nation despite theadversities of war (Magee 1946).

Breast milk is an unequalled way of providingideal food for the healthy growth anddevelopment of infants; it provides all theessential nutrients needed for optimal growthof an infant up to 6 months of age andcontinues to be an important component of thediet of young children between the ages of 6months and 2 years (WHO 2003). Animal milkalso contains the nutrients needed for the

growth and development of their young and istherefore considered to be a valuable food,contributing high quality protein, fat,vitamins and minerals, in the diets of youngchildren, particularly for those that are nolonger breastfed (Dewey 2005).

Milk is an especially good source of Type IInutrients i.e. those that are needed in a fullcomplement for growth and development(Golden 1991). These include all eight of theessential amino acids (that make up a highquality protein), zinc, potassium, sulphur andphosphorous. The high protein quality of milkis particularly important across thedeveloping world where the traditionallycereal based diet of agriculturalists is oftendeficient in some of the essential amino acids.Lysine, for example, is commonly limiting inmaize-based diets. The addition of milk tosuch a diet allows cereal protein to be morefully utilized for growth, by providing themissing essential amino acids. It should benoted however, that the protein in milk willonly be fully utilized for growth where theenergy requirement of an individual is met.Otherwise, some of the protein will be oxidizedto provide energy (Hoppe, Andersen et al.2008). The strongest evidence for theassociation between milk and growth comesfrom observation (Allen, Backstrand et al.1992; Allen 1994; Dewey, Peerson et al. 1995)and intervention studies (Orr 1928; Du, Zhuet al. 2004) that have shown associationsbetween the consumption of cows milk andincreased linear growth in children.

Milk is also a good source of important Type Inutrients i.e. those that prevent specificclinical deficiency diseases and that areimportant for protection against illness(Golden 1991). These include vitamin C (camelmilk particularly), B vitamins such asriboflavin and B6 and calcium. Deficiency of

vitamin A and iodine are two of the threemicronutrient deficiency diseases (MDD) thatare of greatest public health significanceworldwide. Milk is also a good source of thesenutrients.

Fat is important in the diets of infants andyoung children because it provides essentialfatty acids, facilitates absorption of fat soluble


11/50

7

vitamins and enhances dietary energy densityand sensory qualities (Dewey 2005). Wholemilk is a good source of fat and, whenconsumed regularly by a young child underthe age of two years, only a small amount of additional fat (up to 5g/day), is needed to meet

requirements for essential fatty acids.Meeting energy requirements however, ismore difficult with milk alone, and requiresthe addition of carbohydrate-based foods orfats and oils.

Table 2 shows that two cups of milk per daycan meet at least 50% of the recommendedintake of many of the essential nutrientslisted. For young children particularly,nutrient needs per unit body weight are veryhigh and many reviews have shown thatanimal source foods such as milk are essentialif nutrient needs are to be met.

Given the relatively small amounts of foods that are consumed at 6-24 months, the nutrient density (amountof each nutrient per 100 kcal of food) of the diet needs to be very high. Toevaluate the potential for various diets

to meet nutrient needs at this age,linear programming techniques wereapplied to data available from fivedeveloping countries, Peru, Honduras,Guatemala, Ghana, and Bangladesh.The results indicated that a diet based

predominantly on unfortified plant-based foods cannot meet the needs forcertain nutrients at this age,

particularly protein, iron, zinc,calcium and vitamin B12. Therefore, itis advisable to include milk products,as well as meat, poultry, fish or eggs,as often as possible (Dewey, Cohen etal. 2004).

Table 2: Nutritional composition and contribution to recommended nutrient intakes of animal milk

RNI**In250ml*cowmilk

In250ml*goatmilk

In250ml*camelmilk

6-11mths

12-23mths

24-59mths

2 cups milk provide50% RNI or more

Energy (Kcal) 165 173 163 700 900 1500 NoProtein (g) 8 8.9 7.9 13 14 16 YesThiamine (mg) 0.1 0.1 0.1 0.3 0.5 0.6 Yes (younger children)

Riboflavin (mg) 0.4 0.4 0.2 0.4 0.5 0.6 YesNiacin (mg) 0.2 0.7 1.1 4 6 8 No Vitamin B6 (mg) 0.1 0.1 0.1 0.3 0.5 0.6 Yes (younger children) Vitamin B12 (g) 0.9 0.15 0.5 0.7 0.9 1.2 Yes (not goat)Folate (g) 12.5 2.5 1 80 150 200 No

Vitamin C(mg) 3 3 9 30 30 30 Yes (camel only) Vitamin A (g) 95 139 125 190 200 200 YesCalcium (mg) 288 335 317 400 500 600 YesPhosphorous (mg) 300 302 214 275 460 500 YesMagnesium (mg) 30 40 30 54 60 76 YesPotassium (mg) 380 453 354 700 2000 2100 Yes (younger children)Iodine (mg) 0.05 0.06 no data 0.09 0.09 0.09 Yes

Iron (mg) 0.2 0.1 0.5 7.7 4.8 5.3 NoZinc (mg) 1.3 1.4 1.1 0.7 1.7 2 Yes*250ml is equivalent to 1 cup milk. Levels of nutrients in milk taken jointly from Park and Haenlein (2006) andWFP 'NutVal' 2006 v1.4 (Seal 2006)** RNI = Recommended Nutrient Intake - all taken from WHO FAO Vitamin and Mineral Requirements inHuman Nutrition, 2004. RNI = the Estimated Average Requirements + 2SDs, and is therefore sufficient for 95%of the population The RNI for vitamin A gives the EAR only - i.e. estimated average requirement. This is due to the potentialtoxicity of this vitamin when consumed in amounts in excess of individual daily requirements. based on high quality protein


12/50

8

Iron, niacin and folic acid are not wellsupplied by milk. To meet the needs for thesenutrients other animal-source foods areneeded unless multiple micronutrientsupplements or adequate amounts of fortifiedproducts are provided. The amounts included

by Dewey et al in a linear programminganalysis were 50 g egg (1 egg/day) and 14-75g/day of meat, poultry, fish or liver (Dewey,Cohen et al. 2004). Children living inpastoralist populations may have better accessto some of these foods than those living inagrarian communities, where, in most cases,some sort of nutrient supplement or fortifiedfood will be necessary.

As well as providing many of the Type IInutrients essential for growth, milkconsumption also increases bloodconcentrations of insulin-like growth factor I,which is a major determinant of growthduring childhood (Hoppe, Molgaard et al.2006; Olsen, Halldorsson et al. 2007). It hasbeen suggested that the effect of milk onlinear growth may be superior to that of otheranimal source foods (Hoppe, Andersen et al.2008). The milk hypothesis, put forward byBogin, proposes that a greater consumption of milk during infancy and childhood will resultin taller adult stature (Bogin 1998). However,the evidence for this hypothesis is stilldebated.

In pastoralist populations, the transformationof milk into dairy products can help deal withgluts of milk during the wet season andsupply valuable nutrients during the dry.Products such as butter or ghee, yogurt andcheese not only prolong the shelf life of milkbut also change its nutrient profile. Forexample, butter and ghee are a concentratedsource of energy containing approximately 800kcals per 100g, whilst the protein componentof milk has been completely removed in these

products (Platt 1962). Such milk fats are avaluable mechanism for increasing the energydensity of young childrens diets. Whilst thedensity of the fat soluble vitamins such asvitamin A and E is considerably increased inbutter and ghee (there is approximately 2500I.U. of vitamin A in 100g butter and gheecompared to approximately 150 I.U. in 100g of whole cows milk), the water soluble vitamin

content such as vitamin C and the B vitamins,and the minerals such as calcium andphosphorous are reduced to very low levels(Platt 1962; U.S. Department of Agricultureand Agricultural Research Service 2008).

Processing milk into hard cheese alsoincreases the energy and fat soluble vitamindensity of the product by, roughly, a factor of between 5 and 9 (U.S. Department of

Agriculture and Agricultural Research Service2008). Unlike the transformation into milk-fat, transforming milk into hard cheese alsoincreases the density of protein and mineralssuch as calcium (U.S. Department of

Agriculture and Agricultural Research Service2008). Both soft cheese (of the type oftenproduced from milk in the developing world)and yogurt maintain a very similar nutrientprofile to milk (Platt 1962; U.S. Department of

Agriculture and Agricultural Research Service2008).

Many of the components contained in milkhave been shown to demonstrate anti-infectiveproperties. Lactoferrin, one of the proteins inmilk, exhibits both bacteriostatic and

Nomadic woman making charl (souredcamel milk), Turkmenistan


13/50

9

bacteriocidal activity against a range of disease causing microorganisms, and couldimprove immunity against infection inchildren.

Fermented milk products such as yogurt have

been consumed for centuries around the worldand a number of health benefits have beenassociated with them (Heyman 2000).Probiotic (or beneficial) bacteria, eitherpresent in milk or added at the time of fermentation, multiply during thefermentation process. Their growth issupported by milks complex carbohydrates,oligosaccharides (an example of a prebiotic).These carbohydrates will also promote thegrowth of probiotic bacteria in the gut. Suchbacteria have been shown to improveintestinal microbial balance and may decreasethe incidence and reduce the severity of illnesssuch as diarrheas, caused by pathogenicbacteria (Shah 2000).

Summary - milk as a food for children

A high intake of animal milk is linkedto growth in stature and maycontribute to the patterns of childgrowth seen in pastoralist ethnicgroups. It is a very good source of highquality protein and othermicronutrients needed for synthesis of lean body tissue, and increases bloodconcentrations of insulin-like growthfactor.

High quality milk protein allowscereal protein to be more fully utilizedfor growth and development byproviding missing essential aminoacids.

For young children, two glasses(500ml) of milk per day can provide100% of the RNI of protein, riboflavin,vitamin B12 and iodine, and 50% of

the RNI of several othermicronutrients including vitamin A. Vitamin A and iodine are two out of three nutrients that cause the greatestburden of micronutrient deficiencydisease in the world.

The transformation of milk into dairyproducts can help deal with gluts of milk during the wet season and supply

valuable nutrients during the dry.Milk products (butter, ghee and hardcheese) are a particularly good sourceof energy and the fat soluble vitaminscontained in milk. Cheese and yogurtare also a good source of protein and

other micronutrients such as calciumand phosphorous. Iron, niacin and folic acid are not well

supplied by milk and must be providedby other animal source foods, legumes,fortified foods or supplements.

Fermented milk products withprobiotic properties may help tocontrol childhood illness such asdiarrhea.

5. P ASTORALIST RELIANCE ON LIVESTOCKMILK FOR NUTRITION ANDLIVELIHOODS .

The importance of livestock milk to pastoralsurvival was first documented by socialanthropologists in their research among

African pastoralists in the 1950s and 1960s(Stenning 1959; Dupire 1963). The extent of pastoralists dependence on milk and otherlivestock products, and the reasons behindthis, are now well-reported in the literature onpastoralism in Africa. There is by contrast,very little information on this topic among

pastoralists elsewhere.Why pastoralists rely on milk

The answer to why, and the extent to which,pastoralists depend on milk versus otheranimal products has been attributed to anefficiency of energy production, as well as to theavailability of grazing. Milk is available dailyduring certain seasons, meat onlysporadically; and it allows a system whichprovides subsistence for far more people perunit area than any other arid zone productionmethod (Suttie 2001). In parts of eastern

Africa, pastoralists can survive moreexclusively on milk because the double rainfallallows year-round milking. This explains whythe Sahelio-Sudanic pastoralists rely morethan East African herders on agriculture,hunting and fishing (Western and Finch1986).


14/50

10

Sheep in South Sudan

An efficient herder can obtain over 2.5 times the energy from milk and meatthan from meat offtake alone (Western1982).

Livestock herding strategy

Herding multiple species is an adaptivestrategy by pastoralists which reduces riskand extends the milk supplied by livestock.Different livestock species feed on differenttypes of vegetation, thereby maximizingnatural resources. To some extent, losses dueto diseases are also limited by mixed herdsbecause some diseases are species-specific. Indryland Africa, camels are often the mostvaluable livestock species as they possessremarkable abilities to exploit very limitedresources, especially water, and have thelongest lactation period of any livestockspecies (El-Agamy 2006). Sheep and goats arethe most widespread livestock raised bypastoralists world-wide. These smallruminants are particularly important for milkproduction during dry seasons and droughts.

Goats are particularly drought-resistant andcan continue to supply some milk for humanconsumption throughout a dry season. Theymultiply faster than large livestock andproduce more milk through a longer lactationperiod than that of sheep (Degen 2007). Cattle

Afar cattle, Ethiopia

produce relatively high volumes of milk butalso need more water and therefore dry uprelatively quickly during the dry season.

In pastoral and agro-pastoral areas of Africathere is considerable variation in pastoralherd composition by species. In part thisdepends on rainfall and vegetation. As ageneral rule, drier areas support more cameland goat production. These patterns are

clearly evident in countries like Somalia, withcamels and goats found to dominatepastoralism in the dry central rangelands andcattle pastoralism found in the lower Jubaareas in the south (Baumann, Janzen et al.1993). Market forces also affect herdcomposition, and pastoralists adapt tochanging market demands for differentspecies (Al-Najim 1991). Another general ruleis that African pastoralists often associatewealth with ownership of larger species suchas camels and cattle, and poorer householdstend to convert some sheep or goats into cattleand camels when the opportunity arises.

In other parts of the world such as Afghanistan and Tibet, there are comparablepastoralist herding strategies designed tominimize risk and maximize milk supply. Inhigh altitude areas, the selection of livestockspecies includes their ability to withstandsevere cold rather than drought, while also


15/50

11

producing milk. The use of yaks in Tibet is anexample of this strategy, but here again,mixed herds are used comprising yaks andsheep (Nori 2004). In Afghanistan, Kuchipastoralists rear mixed herds of cattle, sheepand goats, with some also keeping camels anddonkeys for transport (Fitzherbert 2007).

In Mongolia, pastoralists:

... herd sheep, horses, cattle, goats,camels, and yaks. Although horses arethe most valued animal, Mongolsactually depend on sheep for theirbasic livelihood as sheep provide milk,which is processed into butter, cheeses,and other dairy products; mutton,wool, and hide for clothes and tents;and dung for cooking and heating.Sheep can be herded on foot, with one

person and a few dogs responsible for a flock (Worden and Savada 1989).

The contribution of milk and milk products toenergy and nutrient requirements

Whilst there are few pastoralist groups inexistence today that exist solely on milk andmeat for their nutritional intake, in the past

many studies have noted the importantcontribution that livestock milk has made toenergy and nutrient requirements. This hasvaried considerably by ethnic group and byseason, with ethnic groups such as theMasaai, Rendille and Turkana of Kenya andTanzania, the Borana of southern Ethiopiaand the pastoralists of Koch county in southSudan seen to consume more than 50% of their dietary energy in the form of milk and

milk products (Galvin 1992; Lindtjrn, Alemu et al. 1993; Galvin, Coppock et al.1994; Homewood 1995; Barasa, Catley etal. 2008). For groups such as the Turkana,studies have noted more than 90% of dietary energy coming from milk during

the rainy season (Galvin, Coppock et al.1994).

Agro-pastoralists such as the AwladHamid in northern Darfur in theSudan, have a diverse diet Camelmilk and milk products only constituteabout 20-25% of the food energy of thediet (Holter, 1988b).

Malian Tuareg (Kel Tamasheq) pastoralist diets include millet andrice, milk and meat, with milk andcereals consumption varying by season(Wagenaar-Brouwer, 1985) .

At the other extreme are the Maasaiand Turkana in Tanzania and Kenyapastoralists , whose diets of milk andother animal products from livestockmay contribute nearly their entirecaloric intake at certain times of the

year (Galvin, 1988; Nestel, 1985).

For the Borana in Southern Ethiopia ...the diet was based mainly on milkand milk products. Children under 5

years of age had, on average, milk for2.3 meals per day. However, the trueamount of milk consumption was

probably higher as young childrenwere often observed drinking milkbetween meals (Lindtjrn, Alemu et al.1993).

Milking Maasai goats, Tanzania

Milking a yak on the Tibet-Qinghai plateau


16/50

12

Whilst many of the studies cited above arenow more than 10 years old, there is somemore recent literature that suggests that milkand animal products in the diets of pastoralistcommunities remain relatively important. Incattle-keeping communities in South Sudan,

milk accounted for up to 30% of total energyintake (Muchomba and Sharp 2006) andamong Nuer agro-pastoralists milk accountedfor 50% of the total energy intake of childrenaged less than 5 years (Fielding, Gullick et al.2000).

The Rendille Sedentarization projectdiscussed below (see section 6) demonstratesthe importance of milk in the diets of pastoralist children in northern Kenyarelative to their settled counterparts (Fratkin,Nathan et al. 2006). At the pastoral householdlevel, milk is a particularly important food forpastoralist children and children have beennoted to receive preferential treatment interms of milk consumption (Barasa, Catley etal. 2008).

In Shinelle, Somali Region Ethiopia,milk is a very important part of the

pastoralist and agro-pastoralist diet,and from a young age children are

given milk. The proportion of childrenover 6 months that had consumed milkduring the previous day was between68% and 94%. This consumption

pattern is related to livestock conditionand migration of animals causing poor

food security at that time (SCUK 2007).

However, a number of factors, includingpopulation growth and a shift fromsubsistence production to commodityproduction, appear to be reducing dependenceon milk and increasing dependence on cerealsas the main staple in some areas.

The Borana (in southern Ethiopia) today are in a state of considerablechange that has been induced

primarily by a long-term decline in the per capita supply of cow's milk, thetraditional dietary staple. Thisimbalance has resulted from steady

growth in the human population in

combination with density dependent fluctuations in cattle population(Coppock 1993).

Over the past 20 years there have been anumber of studies of pastoral production

systems that have clearly shown that, whilstanimal products remain important, it wouldbe difficult for pastoralists to subsist withoutgrains and sugar and that these commoditieswere purchased in increasing amounts duringthe last decades. This shift is attributed to adecrease in livestock productivity and a shiftfrom a subsistence milk-surplus system to asale or commercial live animal system(Abdullahi 1993; Fratkin 2001).

Hard choices: milk for calves, milk to eat,milk to sell or milk for social exchange

Milk from livestock can be used in severalways to support a pastoral family and thisoften means hard choices, especially for poorerfamilies.

Traditional subsistence pastoralists inEast Africa tend to keep large herds,milk cattle in preference to eating them, and subject them to long

foraging treks.

Milk must be available in sufficientquantities to sustain calves to weaning age. The surplus is available forhuman consumption. Offtake thatdeprives calves, though benefiting the

pastoralist in the short-term, willimpair herd replacement and growthrate in the long-term. The herder musttherefore balance his immediate andlong-term needs when milking.Moreover, because milk yields declinethrough the dry season for both the calf and human dependent, the herder

must rely increasingly on alternativelivestock products, or other foodsources (Western and Finch 1986).

The importance of preserving the herd inpastoralist communities for future food andlivelihood security and social status and tieswas noted by de Waal in his eminent text onFamine that Kills in Darfur, Sudan.


17/50

13

Pastoralists may even, underconditions of stress, endure serioushunger to sustain the household herd (de Waal 1989).

This phenomenon has been noted widely

within the literature on the copingmechanisms that communities employ inresponse to food insecurity (Corbett 1988;Maxwell 1996).

The exchange of milk and milk products forincome and more energy dense grains or otherstarchy food is an important pastoral strategyto fulfill caloric needs, and is practiced widely.In Borana, southern Ethiopia for example,grain purchased with the sale of livestockproducts met over 30% of dietary energyrequirement (Cossins and Upton 1987) and inSomali Region, Ethiopia milk production wasnoted as the most important source of revenue(66% of total income) (Barrs 2000). Sales peakin the wet season and both wealthierhouseholds and households closer to themarket have been noted to sell or exchangeconsiderably more (in quantity) than others(Holden and Coppock 1992). However, poorerhouseholds have been observed to takeproportionally higher milk off-take per animalin order to purchase grains for humanconsumption (Holden, Coppock et al. 1991).This has been observed by livestock scientiststo lead to a cycle of poverty: ill-fed younganimals mature more slowly to reproductiveage (so the herd or flock does not expand asfast) and are weaker and more susceptible todrought, cold and disease. Their humanowners, however, are able to provide a littlemilk to their families, including children.

Pastoralists have other choices too. One of these is whether to convert some of the rawmilk into some form of preserved dairyproduct such as yogurt, cheese, butter or ghee

(separated butter fat). Pastoralists mustdecide whether to exchange some of theseproducts for other goods, or retain thepreserved form of protein and fat as a foodsource for the family in lean periods. InSomali Region, Ethiopia a high proportion of cattle milk is converted to ghee (SCUK 2007)for consumption or sale, with wealthier

households converting substantially morethan poorer households.

The social value of milk and dairy productsalso plays an important role in decisionmaking. Rather than sell or consume surplus

milk products, pastoralists often use lactatinganimals or their milk products to maintainsocial ties. In some pastoral societies, thereare cultural prohibitions on selling milkproducts, which must be kept for hospitalityand sharing between group members. Thesepractices are well-documented for manypastoral African societies (Oba 1994; Bush1995) and also occur in Afghanistan(Fitzherbert 2007). When decisions must bemade on how to allocate milk from animals,the future social insurance gained by sharingmilk between group members may outweighan immediate financial gain from its sale.Likewise, children of pastoralists often haveinformal access to livestock milk off-take fromother families in their social group, makingmeasurement of childrens milk consumptionsomewhat problematic.

Impact of commercialization on the use of milk

The pan-African economic review published bySikana, Kerven et al. (1993) is probably themost significant commentary to date on theimpact of commercial change in pastoral

Africa. It focuses on the shift in the value of milk, which is influenced by a number of discrete and identifiable variables.Importantly, it documents the diminished roleof milk in pastoral diets undercommercialization, especially among poorerhouseholds. Better-off households with acalorically sufficient diet can afford to viewmilk as a source of protein, while caloricallyinsufficient households view milk as a sourceof calories and, as a result, practice higher

rates of milk offtake for sale to purchasecereals for consumption. Other studies notethat the closer (poorer) households are tomarkets, the more milk offtake is taken forsale. Because of this greater dependence onmilk income, the management strategies andmilking practices of poorer pastoralists differfrom those of richer households. They oftenlive nearer to watering points, despite range


18/50

14

being of poorer quality there, and wateranimals more frequently as a strategy tomaximize milk output.

The greater dependence of poorerhouseholds on dairying income may

also reduce their herd mobility, which ,in turn has been linked to milk output[through reduced access to quality

grazing (Sikana, Kerven et al. 1993).

These findings together with calf morbidity and mortality data lendsome support to a causal relationamong factors of proximity to market,dairy sales and risk to calf health for

poorer households. Dairy marketing seems to be an appropriate survivalstrategy for poorer households,however it's associated short and long term risks (cuts in milk productionand calf mortality and problems inhuman nutrition) should be reduced byappropriate interventions that createalternative income opportunities andimprove calf feeding management(Holden, Coppock et al. 1991).

The role of milk in the diets of infants andyoung children in pastoralist settings

Several studies (in eastern Africa) thatexamine the dietary intake of infants andyoung children in pastoralist areas identifysome common themes with regard to the useof animal source foods. These include veryearly introduction (often within the first fewdays of life) of animal milk (sometimes mixedwith water) and/or an animal fat (such asghee) to supplement breast milk. Intake of these complementary foods continues, alongwith breastfeeding, throughout the first sixmonths of life and beyond, when milk is oftenmixed with cereals to make porridge. The two

most substantial bodies of research recentlygenerated on the role of milk in infant andyoung child feeding in pastoral communitiesconcern the Datoga of Tanzania (Sellen 1998;Sellen 2001; Sellen 2002) and the Turkana of northern Kenya (Gray 1994; Gray 1995; Gray1996; Gray 1996; Gray 1998).

Daniel Sellens longitudinal studies on theDatoga found that supplementary foods fed toinfants and young children included fresh andsoured goat or cows milk, maize-meal grueland the principal adult staple, stiff maize-meal porridge; that the median ages of

introduction of non-human milks, introductionof grain-based solids, and cessation of breast-feeding were 2, 11 and 20 months respectively;and that feeding practices show a lack of concordance with international standards.

Sandra Grays important study on butterfatfeeding in early infancy (Gray 1998) builds onearlier research among Turkana pastoralists

(Gray 1994; Gray 1996) and found similarpatterns of breastfeeding and complementaryfeeding as among the Datoga (including earlysupplementation of breast milk with animalmilk and with butterfat).

Save the Children UK, in their report on thecauses of malnutrition in Somali RegionEthiopia, reported that only 6.6% of infantswere exclusively breastfed, mostly due to thecommon practice of giving animal milk early.

[In Somali Region] From a young agechildren are given milk. In additionto breast milk children less than 6 months consumed animal milk. Up to12 months the children still consumed

predominantly animal milk withapproximately 20 % consuming cerealand 14% oil. The most common milk is

goats followed by cows. Cows milk is

Baggara Arab nomad with milk containers, Darfur


19/50

15

more common in agro-pastoralcompared to pastoral.

Lindtjorn, in his work with the Boran insouthern Ethiopia also reported the practice of supplementing breast milk with cows milk

from the first few days of life. Understandingthese practices and their consequences fornutritional status is further discussed undersection 6 below.

Summary - pastoralist reliance on livestockmilk for nutrition and livelihoods

Herding multiple species is anadaptive strategy by pastoralistswhich reduces risk and extends themilk supplied by livestock. Whilstcamels are the most valuable of alllivestock species, goats and sheep aremore common and can continue tosupply a little milk for humanconsumption through the dry seasonwhen cows cease lactating.

During the 80s and 90s manypastoralist ethnic groups in eastern Africa were found to consume between50% 90% of their dietary energy asmilk and milk products. Although thisproportion may now be changing, therelative importance of milk andanimal products in the diets of thesegroups seems likely to remain.

Exchange of milk and animal productsfor grain is an important strategy tofulfill caloric needs for all pastoralists.

The balance between milk offtake forhuman consumption and/or exchangeand maintaining calf health is adifficult one, especially for poorerfamilies for whom total milk yields arerarely sufficient to cover consumptionrequirements. Choices may meanputting human health at risk to

preserve animal health. Commercialization (producing animalproducts for market) is reducingdependence on milk and increasingdependence on cereals as the mainstaple in some areas. This is especiallytrue for poorer households nearer tomarkets, who convert a higherproportion of milk to higher calorie

grains and, to maximize milk yields,may use herd management strategiesthat puts calf health at risk.

Sharing of milk within a social groupis a strong cultural practice for mostpastoral communities. It improves

access to milk for children within thesame group as well as acting as animportant social insurance strategy inleaner times.

Contradictory to internationalrecommendations milk and/or ananimal fat (such as ghee) are oftenintroduced into the diets of infantsfrom a very early age in pastoralistcommunities. They continue to be themost common complementary foodsgiven to infants and young childrenthroughout the first six months of lifeand beyond, with milk often beingmixed with cereals to make a porridge.

6. H OW P ASTORALIST DIETS ANDLIVELIHOODS AFFECT NUTRITIONALS TATUS

Diets and nutrient intakes

This literature review highlights theconsiderable inter-ethnic group variability indietary intake and intra-ethnic group

variability by season. However, there are anumber of common themes that stand out inthe literature as being importantdeterminants of the nutritional status of pastoralist people, particularly of women andchildren.

Reliance on milk: a protective or harmful factor? In contrast to the diets of most populationsacross Sub Saharan Africa, pastoralistsgenerally have good access to high qualityprotein and other Type II nutrients in theform of milk and, seasonally, blood and meat.Several studies, starting with those led byGalvin in the 1980s, noted that these pastoraldiets were adequate in terms of protein(providing between 200-400% of daily proteinrecommended intake in many cases) but werelow in energy. This became known as theprotein-rich calorie-poor phenomenon (Galvin


20/50

16

1985; Galvin 1992; Homewood 1995; Fratkin,Roth et al. 1999; Fratkin 2001).

As discussed under section 3, researchershave suggested that it is the high level of goodquality protein intake and other Type II

nutrients that, in part, results in thesignificantly lower levels of stunting foundamong pastoralist people, compared to theiragrarian neighbors. High levels of Insulin-likeGrowth Factors (IGF) found in milk are alsolikely to play a role in this. However, this linkis still debated.

Fratkin and others have noted that the mainnutritional challenge for pastoralists isobtaining sufficient calories, and that mostpastoralists supplement their livestock-baseddiets with grains and sugar to providecarbohydrate calories and to maintain foodsupply during the dry season and droughtwhen milk yields fall. Several studies havequantified inadequate levels of energyprovided by pastoralist diets reliant on milkand animal source foods and have linked thisto high levels of wasting (measured by weightfor height). Homewood, in her studies of theKenyan and Tanzanian Maasai identified highlevels of wasting and noted populationssurviving on 70% of internationalrecommended energy intakes (Homewood1995). Several authors have found energyintakes of pastoralists in northern Kenyaranging from 1000-1400 kcal per person perday, only about 50% of the average daily adultrequirement, and have linked this to the lowbody mass indices identified (Little 1989;Galvin 1992; Galvin, Coppock et al. 1994).However, none of these studies accounted forthe different body shapes of pastoralist peoplewhen using nutritional indices like body massindex and weight for height. This may havebiased their assessment of nutritional status.

Increasing milk intake is protectiveagainst wastingIt is important to note that milk (particularlywhere intake is high) is an importantcontributor to total nutrient intake in manypastoralist populations. Therefore, it followsthat increasing intake of milk (which providesenergy as well as protein) is protective againstwasting, as well as stunting, in young

children. This is shown with multipleregression analyses in a number of reportsand studies (see discussion on sedentarizationbelow) (Nathan, Fratkin et al. 1996; Fratkin,Roth et al. 2004; Grobler-Tanner 2006; SCUK 2007).

Vitamin and mineral deficiencies arerare apart from iron deficiencyanaemia which is common

The role of animal source foods (ASF), thatinclude milk, meat, eggs and blood, inimproving micronutrient status is well known,in particular for vitamin A, vitamin B12,riboflavin, vitamin C (camel milk only),calcium, iron and zinc levels (Bwibo andNeumann 2003; Murphy and Allen 2003;Neumann, Bwibo et al. 2003).

Milk alone is a good source of many of thesemicronutrients and, as a result, some authorshave noted micronutrient deficiencies as beingrare among pastoralist populations thatconsume large amounts of milk along withother foods (Fratkin, Roth et al. 1999).However, iron, folic acid and niacin areprovided in very small amounts in milk, andtherefore pastoralist populations may be atrisk of deficiency of these micronutrients.Deficiencies in iron and folic acid causeanaemia, and this has been commonly notedin studies of pastoralist populations. About25% of Maasai women and children studied byNestel et al were anaemic (Nestel andGeissler 1986) whilst among the Rendillepastoralists of northern Kenya, Fratkin et alalso noted the prevalence of iron deficiencyanaemia as being high (although not as highas their settled counterparts) (Fratkin, Rothet al. 1999). However, there are alsoimportant non-nutritional risk factorsassociated with anaemia; malaria andhookworm are the most important of these,and both are common in many pastoralist

settings (Schelling, Daoud et al. 2005; SCUK 2007).

Low levels of fruit and vegetable intake havealso been noted in some pastoralistpopulations which may put them at risk of vitamin A and vitamin C deficiencies wheremilk (particularly camel milk) is notconsumed in sufficient quantities to meet


21/50

17

requirements (Zinsstag, Schelling et al. 2002).This is discussed further below.

SCUK, in their recent analysis of thepastoralist diet in Somali region, Ethiopiasuggested that the main risks for

micronutrient deficiencies were from iron andfolic acid, making anaemia a particular risk.Zinc and vitamin C were consumed atmarginal levels, although the analysis did notaccount for any wild food consumption,thought to make some contribution to intakeof these nutrients (SCUK 2007).

Supplementing milk with a range of other foods is important

Although studies do support the use of animalmilk as an important source of manynutrients some have also noted theimportance of diet diversity to ensure thatadequate levels of all nutrients are met.

This review has already noted the importanceof supplementing animal foods withcarbohydrate-based foods if pastoralists are tomeet their energy requirements. It is worthnoting here the particular importance of sugar. Several studies and analyses havenoted the high consumption of sugar amongpastoralist populations; it is considered acrucial addition to cereal-based mixes forchildren and to tea for the whole family bymany pastoralist groups (Abdullahi 1993;FSAU and FAO 2004). The Household FoodEconomy Groups livelihood zone analyses andSCUKs analysis of Somali caring practicesfound it to be a prioritized expense across alllivelihood zones with consumption risingwhen availability of milk fell (2008). It islikely that sugar plays an important role infilling the energy gap (FSAU and FAO 2004).

Human serum retinol and livestock milkretinol levels were assessed as part of a study

on the health status of Chadian nomadicpastoralists and their livestock. Of theexamined women (n = 99), 43% weremoderately retinol-deficient and 17% severelyretinol-deficient. None of the interviewedwomen (n = 87) reported the consumption of fruit, and only two reported consuming freshvegetables in the past 24 hours. Milk is thealmost exclusive source of vitamin A for these

populations and vitamin A intake was closelycorrelated to retinol levels in milk. Zinsstagconcluded that although milk was animportant source of retinol, intakes from milkneed to be complemented by green leafyvegetables, fruits, and supplements (Zinsstag,

Schelling et al. 2002).This conclusion is also highlighted by the FoodSecurity Analysis Unit (FSAU) in their studyon Infant and Young Child Feeding andHealth Seeking Practices in Somalia. Theyconclude that diet diversity is associated withchanges in levels of wasting acrosspopulations. This underlines that higherconsumption of all foods, and ideally a varietyof foods, supports good nutritional status(FSAU 2007). This association has beenwidely noted in populations across thedeveloping world (Ruel 2002).

The effect of season on nutrient intakeand nutritional status

The considerable variation in milk supply byseason for many pastoralist groups isdiscussed under section 3.

Recovery of nutritional status afterdry season and drought

A common conclusion in the literaturereviewed is that the nutritional status of pastoralists appears to be relatively wellprotected against short-term declines in foodavailability i.e. during seasonal changes inmilk and food availability, and recovers morequickly after a dry season than their agrarianneighbors. In southern Ethiopia, Lindtjornobserved a more rapid improvement innutritional status of pastoralist Boranchildren after the start of rains compared tochildren living in a Boran agriculturalcommunity. He found that outside of droughtyears, the pastoral livelihood is more resilientto seasonal variations as milk yields increase

much more quickly than food production fromcereals once rains arrive (Lindtjrn, Alemu etal. 1993). Studies among the Ariaal, Rendille,Dagota and Turkana find that despitesignificant seasonal variation in rainfall andfood availability, child nutritional statusshows only moderate, if any, decline in the dryseason (Shell-Duncan 1995; Sellen 2000).Other studies describing patterns of wasting


22/50

18

in Somalia and northern Kenya note that thepeak in levels of wasting (in normal years)does not necessarily occur in the traditionalcereal-based hungry season suggesting thatpastoralists might be relatively protected fromthe usual cereal gap (Chotard, Mason et al.

2006).Some reports suggest that those groups thatkeep higher numbers of camels and havemaintained strong social networks thatemphasize food sharing (for example the Ariaal and Rendille in northern Kenya) aremore protected in drought than those groupsthat rely more on cows and small livestockand that do not have such a strong culture of social protection (e.g. those in south Turkanaand some Somali ethnic groups) (Roth,Fratkin et al. 2003).

Whilst pastoralists may show relativeresilience against normal seasonal changes infood availability, some reports have observedthat nutritional status decline is considerablymore marked as a result of longer term foodinsecurity such as successive droughts over anumber of years. Chotard et al noted that forthe pastoralist populations of Somalia andnorthern Kenya nutritional data analyzed byseason indicated that fluctuations in levels of wasting were about five percentage points innormal years with considerably larger shiftsseen in years of subsequent drought (Chotard,Mason et al. 2006).

This seeming resistance to drought butvulnerability to destitution has long beendiscussed in the literature:

If they are able fully to employ theirrange of coping strategies, of whichmobility is central, pastoralists areresilient to drought [Ellis, 1985].Turkana herders, for example, cansurvive a failed rainy season or two

without external assistance.Nonetheless, not all pastoralists areresilient. Their degree of vulnerabilitydepends on several factors: access todry season ranges; peace withneighboring pastoral communities; theratio of female to male stock; and theavailability of, and access to,commercial cereals [Swift, 1989]. Herd

sizes also determine vulnerability. The greater the buffer stock of animals, the greater the likelihood that householdscan weather a drought on their own.The real Achilles heel of pastoralism isthe slow pace of recovery after a

devastating drought. Just as the slideinto destitution is a slow spiraldownwards with total collapseoccurring very rapidly at the end, sorecovery is a long, uncertain journeyback up. It takes several seasons forherds to rebuild to a point of reasonable food security. Until thisthreshold is reached, milk will bescarce, and the sale of stock to obtainmaize, while necessary, will be limited(Bush 1995).

The effect of animal disease on milkavailability and nutritional status

Some livestock diseases cause the death of animals and so directly and permanently stopthe supply of milk from affected adult females,and indirectly prevent purchases of milkbecause fewer animals are available for sale.Other livestock diseases affect productivityand therefore reduce the volume of milk whichis available for direct consumption, processingor sale. Given the high importance of milk inthe diet of pastoralists (see section 5) it hasbeen assumed that livestock disease has adirect and major impact on household foodsecurity. There has been little mention in theliterature however, of the direct link betweenreduced milk yields due to animal disease andthe impact that this might have on thenutritional status of pastoralist children. Nostudies were identified that have examinedthe causal relationship between the two.

Conventional studies and research The assumption that livestock disease has a

major impact on pastoral household foodsecurity seems reasonable given that standardveterinary textbooks note that variousdiseases cause reduced milk production,reduced fertility or other losses which affectmilk supply. As an example, chronictrypanosomosis affects cattle and camels inmany pastoral areas of east Africa and is welldescribed in texts (Stephen 1986; Brown,


23/50

19

Hunter et al. 1990; Radostits, Blood et al.1994).

Although there is a reasonable body of research which quantifies milk production indifferent livestock species in pastoral areas of

Africa (Wilson and Clarke 1976; Mefit-Babtie1983; Simpkin 1985; Oxfam and Ireland 1990;Nauheimer 1993; Kaufmann 1998; Zeleke andBekele 2001) the losses due to animal diseaseare rarely quantified in the veterinaryliterature.

The limited quantitative data on milk lossesdue to specific diseases might be explained, atleast in part, by the approaches and methods

of conventional livestock research. Forexample, the measurement of milk productionrequires summation of the milk which isconsumed by offspring plus milk which isobtained during milking for human use. Theformer is difficult to measure and requiresaccurate weighing of calves, lambs or kidsimmediately before and after feeding; thevolume of milk consumed is then calculatedfrom the difference between these two weightmeasurements. This methodology iscumbersome, especially when large scales areneeded to weigh calves, and when pastoralherds are moving. A further complication of milk measurements in pastoral herds is adhoc milking by herders, especially children,while animals are grazing. This off-take awayfrom the homestead is very difficult tomeasure. In addition, factors including theavailability and quality of vegetation, watersupply and distances traveled to grazing areasor water sources all influence milk production.

Therefore, studies on disease impacts need tounderstand the relative importance of thedisease in question compared with these otherfactors.

Pastoralist views

Alternative perspectives on the impact of livestock diseases on milk supply are providedby pastoralists themselves, and include bothqualitative and quantitative estimates. Forexample, the use of participatory methods hasenabled pastoralists to define their ownindicators of disease impact and comparedifferent diseases. In Somaliland the indicatorDisease reduces milk yield was one of 26indicators identified by herders whencomparing livestock diseases (Mohammed1993), and pastoralists in South Sudan(Catley, Okoth et al. 2001; Barasa, Catley etal. 2008), Tanzania (Catley, Chibunda et al.2004) and Ethiopia (Rufael, Catley et al. 2008)have all included a milk loss indicator in theirdescriptions of important livestock diseases.

Using participatory methods in South Sudan,Nuer herders estimated milk losses in cattledue to acute foot-and-mouth disease at 1.6liters per day (95% CI 1.20, 1.95 liters perday) for an average of 14 days; thisrepresented a 62% fall in production inaffected cows during the period of clinicaldisease (Barasa, Catley et al. 2008).By combining this information with estimatesof disease incidence by age group, the reducedvolume of milk available for humanconsumption due to foot-and-mouth disease(FMD) was estimated at 839 liters perhousehold. Assuming intake of no other foods,

Trypanosomosis causing emaciation and reducedmilk production in cattle, Somali region of Ethiopia

Research in South Sudan on the livelihoods andnutritional impact of foot-and-mouth disease


24/50

20

this volume of milk could feed one adult for240 days or a family of two adults and fivechildren (between two and 12 years of age) for47 days. The study also explored theseimpacts in terms of seasonality in humandiets, with high dependence on milk at certain

times of year. Outbreaks of FMD occurredshortly before the hunger gap period in SouthSudan, when milk was a particularlyimportant component of the diet.

These participatory impact assessments showthat pastoralists associate specific livestockdiseases with reduced milk production, andcan rank or score diseases according to theimpact on milk supply. Similarly, pastoralistscan assess the value of disease control againstindicators such as milk production, and canestimate absolute reductions in milk off-takeper milking by disease. Pastoralists are wellaware of the multi-factorial nature of milkproduction, and can describe the relativeimportance of factors such as rainfall, pastureand conflict for milk supply compared withlivestock diseases.

Understanding infant and young childfeeding practice and its consequences

In FGDs women were asked what foods they gave to an infant in the first40 days. Water with sugar was almostuniversally provided and animal milkand 'badhi' (fatty tissue from sheep).When asked why they thought thatbreast milk is not sufficient for a newbaby most replied that more vitaminswere available from the other foods(SCUK 2007).

The important role that milk and animal foodsplay in the diet of children from a very youngage is discussed above under section 5. Thetiming of the introduction of complementary

foods has been widely noted in the literatureas not complying with international standardsin pastoralist populations. Animal milk iscommonly introduced from a few days old andit is suggested that this may reduce thefrequency of breastfeeding. This hascommonly been cited as a risk factor formalnutrition among pastoralist infants underthe age of six months (Sellen 1998; Sellen

2001; SCUK 2004; Nyaruhucha, Msuya et al.2006; SCUK 2007) although none of thesestudies directly measured wasting in thisgroup.

There are other risks associated with this

practice. Not only is the iron content low andpoorly absorbed in animal milk but its earlyintroduction to infants less than nine monthshas been shown to provoke microscopicintestinal bleeding. This could increase therisk of poor iron status and anaemia ininfants, although a causal link betweenintestinal blood loss and iron status has notbeen demonstrated (Ziegler, Jiang et al. 1999;

Agostoni, Decsi et al. 2008). In addition therenal solute load of animal milk is high due tothe relatively large amount of protein andseveral minerals (sodium, chloride, potassiumand phosphorus). This could lead tohypernatremic dehydration and is a particularrisk for infants under six months of age thatreceive all of their nutrients from a singlesource (breast milk or formula) and do notconsume water (Dewey 2005).

The introduction of complementary foodsother than milk to young children inpastoralist societies is often late (at ninemonths or older) and of poor diversity. Thishas also been noted as a risk factor formalnutrition in young children particularlybecause, on average, rates of wasting inpastoralist populations tends to be lower, incomparison to settled populations andnational averages, below six months of age,but rises quickly after six months (Little, Grayet al. 1993; Gray 1996; Sellen 1998; Sellen2001; Sellen 2002; FSAU 2007).

In FGD, when asked why they do notstart (weaning) earlier, the pastoraland agro-pastoral women gave the

following reasons: there is food

shortage so they give just milk; thechild cannot eat sorghum which is themain food available; the child lacksteeth for chewing; complementary

foods are not needed; because the childwill develop diarrhoea; breast milk issufficient. To enable earlierintroduction of weaning foods, theseconstraints will need to be addressed,


25/50

21

i.e. both knowledge and practicalconstraints of food availability (SCUK 2007).

However, it is widely agreed that milk andanimal food consumption that complements

breast milk for children over the age of sixmonths is protective against malnutrition(SCUK 2007; Dewey and Adu-Afarwuah2008). The literature that examines the effectof settling on the nutritional status of pastoralist children demonstrates this well.This is discussed below.

The reasons behind early introduction of animal foods in the diets of infants andreduction in frequency of breastfeeding havebeen widely debated in the literature. Gray, inher studies of the Turkana in the 90s,explains these practices as adaptive responsesto a highly stressful reproductive environmentcharacterized by high maternal and infantmorbidity and mortality and chronicundernutrition (Gray 1994; Gray 1996; Gray1998). The structure of the pastoralist systemrequires high human fertility to maintain thefamily herds at adequate levels. Theirchildbearing role requires that Turkanawomen balance the needs of the childcurrently breastfeeding with the demand forlarge families. The implications of this aresuggested by the finding that the real reasonfor early introduction of complementary foodsis to shorten birth intervals, a result echoedby Sellen (Sellen 2001).

Others have found that Datoga and Turkanabirth spacing strategies have a stronglyseasonal component, with mothers preferringto terminate breastfeeding at the end of thelong rains before food supply becomes scarceand their own nutritional status deteriorates(Leslie and Fry 1989; Sellen 2001). Labordemands on the lactating mother have also

been linked to a reduction in breastfeedingfrequency. The high workloads of the pastoralsystem, and work that requires considerabletravel, exert significant pressure on the timeavailable for feeding (Gray 1994; SCUK 2007).

Finally, some have suggested that the earlyintroduction of animal milk into the diets of young infants is due to low breast milk

volume (Ettyang, van Marken Lichtenbelt etal. 2005) and that the energetic benefitsaccrued to nurslings by increased intake of foods such as butterfat may be of greatersignificance than any reduction in exposure toinfection (Gray 1998). Several studies have

shown that, at least in the first six months,breast milk production can be maintained atlevels sufficient to satisfy infant needs and istypically maintained even when mothers aremoderately undernourished (Cohen, Brown etal. 1994; Prentice and Goldberg 2000). Due tothis work it is generally thought that theinfant is reasonably well protected againstany deterioration in nutritional status of themother. However, this work does not take intoaccount womens own perception of how ablethey are to provide sufficient nutrition fortheir infants. This is an important factor toconsider when examining factors associatedwith the frequency and duration of breastfeeding (Gray 1996).

The links between maternal and childnutritional status

It is well established that maternalnutritional status is an important predictor of the nutritional status of her young children.The first article in the recent LancetNutrition Series discusses the strongassociations between maternal short statureand cephalopelvic disproportion (i.e. the babyshead being proportionately too large for safedelivery) and between low maternal body-mass index and intrauterine growthrestriction (Black, Allen et al. 2008). Severeanaemia has also been linked to the risk of premature delivery and low birth weight(Huma, Salim Ur et al. 2007).There is only one study in pastoralistpopulations that has linked maternalnutritional status to birth weight. Pike foundthat infant birth weight was correlated to the

seasonal pattern of maternal exposure toillness and fluctuation in nutritional statusamong the Turkana in northern Kenya (Pike2000). The Rendille Sedentarization Projectfound that maternal dietary intake andnutritional status was highly seasonal inpastoral communities (although better thantheir settled counterparts) (Fujita, Roth et al.2004; Shell-Duncan and Yung 2004) and


26/50

22

anaemia, whether through dietary or infectionpathways, has been noted as prevalent insome pastoralist populations (see section 6).However, none of these studies examined thelink with infant birth weight nor were theyable to comment on the prevalence of low birth

weight in the populations under study.Maternal nutritional status (measured by midupper arm circumference) was found to be oneof the factors independently associated withchild wasting and stunting in the Somaliregion of Ethiopia (SCUK 2007). Thisrelationship may be working through anumber of pathways including direct maternalnutrition-infant nutrition links, maternalworkload and caring practice.

Black et al. (2008) reaffirm that maternalunder-nutrition has little effect on the volumeor composition of breast milk unlessmalnutrition is severe. There is evidencehowever that demonstrates the importance of maternal nutritional status for theconcentration of some micronutrients(particularly vitamin A, iodine, thiamin,riboflavin, pyridoxine, and cobalamin) inbreast milk. The risk for infant deficiency of these micronutrients is increased by maternaldeficiency. Although these micronutrientdeficiencies have not been commonly noted inpastoral populations, low human serumretinol has been demonstrated in nomadicpopulations in Chad (see section 6) (Zinsstag,Schelling et al. 2002). However, this study didnot examine vitamin A levels in breast milknor risks to infant nutritional status. Theredoes seem to be a link between mothers ownfeeling of well being and the frequency andduration of breastfeeding (see section 6)although a direct association with poornutritional status among infants has not beendemonstrated.

Sedentarization and commercialization

There is a global trend for mobile or nomadicpastoralists to cease moving, either partiallyor completely settling with their animals. Thistrend is sometimes linked to increasedpastoralist participation in markets, andsometimes results from impoverishment, lossof grazing land, climatic disasters and othercauses, all of which are now quite well-

documented.

As a result of settlement, some work hasdocumented an improvement in nutritionalstatus among children and adults due to anincrease in commercial opportunities,

improvements in child care practices andbetter access to health care. Ensmingersstudy of the economic transformation of theOrma of Kenya for example found increasedresidence in market centers and agriculturalcommercialization associated with improvednutritional markers (weight for height) foradults and male children but not for femalechildren (Ensminger 1991).

However, whereas settlement may hold someopportunities, these can be at the expense of herd productivity, and in particular, milkyields. A comparative study of milk yieldscarried out among the Ngok Dinka of Sudanindicates that milk production in migratoryherds is about 50% higher than in sedentaryherds during both the wet season and the dryseason (Niamir 1982).

By far the largest work that has examined theconsequences for child nutrition of sedentarization is that of the RendilleSedentarization Project in northern Kenya,which ran for several years in the mid-90s. Bycomparing the diet, nutritional status, fertilityand morbidity of women and children innomadic and sedentary communities theywere able to show, quite conclusively, thatchild nutrition and growth patterns wereworse in the sedentary communities incomparison with the nomadic control group(Fratkin, Roth et al. 1999). What clearlydifferentiated the nomadic from the sedentarycommunities was the consumption of milk.Based on 24-hour dietary recall the projectwas able to show that children in Lewogoso(the nomadic community) consumed on

average over ten times as much milk (3.2 vs.0.3 cups) as children in Songa (the settledcommunity). It was particularly during thedry season that the sedentary communitieshad significantly more malnutrition andsignificantly higher levels of anemia andinfection such as diarrhea. This studysuggests that the pastoral nomadic diet,particularly one dependent on camels' milk,


27/50

23

offers children better resistance to thepressures of drought and supports findingsthat the subsistence base of mixed-speciespastoralism is superior to sedentaryalternatives with respect to child health(Nathan, Fratkin et al. 1996).

Other studies have noted similar findings.Children of settled Bedouins had a poorernutritional status than both the true nomadsliving under the same environmentalconditions and the semi-settled agro-pastoralpopulations (Baba, Shaar et al. 1994). In 2006Grobler-Tanner found that pure pastoralistswith animals in Turkana were comparativelybetter off than those who had settled and weredependent on food aid, trade and begging inurban and peri-urban areas (Grobler-Tanner2006).

Wealth

The influence of economic status on dietaryintake and nutritional status of pastoralistpeople is debated in the literature. Studies onBorana, Turkana, Rendille and Datogapastoralists in southern Ethiopia, Kenya andTanzania examine the interactions of marketaccess and household wealth on nutritionstatus. Several of these studies find wealth tobe a poor determinant of nutritional status inthese populations. They highlight the positiverole that intra-household dynamics and thesharing of resources amongst social groupshave on nutritional status during periods of decreased food availability (Galvin 1992;Galvin, Coppock et al. 1994; Homewood 1995;Fratkin, Roth et al. 1999; Shell-Duncan andObungu Obiero 2000; Sellen 2003). Between2000-2006, the nutritional status of childrenin Somali Region, Ethiopia did not changewith wealth (Mason, Chotard et al. 2008).

This study finds no significant

differences in diet and nutritionalstatus between Maasai communities,and other studies have failed

milk matters: a literature review of pastoralist nutrition and programming responses

Documents