MORBIDITY, MORTALITY AND DIET AS INDICATORS OF PHYSICAL AND ECONOMIC ADAPTATION AMONG BOLIVIAN MIGRANTS CONNIE
Department
WEn.
of Geography. University oJ Minnesota Miuneapolis~ MN 55455, U.S.A.
Abstract-Bolivian peasants are migrating from marginal farming areas in the highlands to newly accessible settlement areas in the tropical lowlands. Local census health data and a community nutrition survey indicate that the migrants have a better diet and participate more in the market economy than do peasants in the highlands, but probably at the expense of a higher toddler mortality rate among their children.
farming areas in many low-income countries are characterized by excessively small landholdings, underemployment, and poverty. One result is massive rural-urban migration. Yet the cities are illprepared to absorb the new arrivals, and most of the migrants are ill-prepared to compete for the few jobs available. The cities are characterized by crowding, unemployment, and the poverty of slums. Several of the low-income countries in Latin America and other parts of the world have extensive sparsely settled areas, especially humid tropical lowlands. As new highway construction links these to populated areas, people migrate to establish farms there. This provides at least some temporary relief of the tensions resulting from population growth under conditions of limited access to land or jobs in the traditional farming areas and the cities. Many studies have attempted to evaluate the success, or adaptiveness, of resettlement by small farmers in tropical lowlands. Most of these have evaluated government-directed migration and settlement, although throughout Latin America spontaneous settlement accounts for the great majority of migrants. Most studies evaluate the success of new settlement in strictly economic terms, usually by whether the government investment in a project pays off in terms of income generated by increased agricultural production [l J. This is a valid way to evaluate projects, but rt addresses only a small part of the settlement process. Increased agricultural production in new settlement areas is frequently the result of large-scale commercial agricultural enterprises which benefit more than do small farmers from infrastructure such as new roads, electrification and technical assistance. In addition, the goals of small farmers and governments are not necessarily the same, and the short-term and longterm goals of either or both often are not congruent. Given these complexities, defining “success” in a new settlement area is difficult; finding data to evaluate any aspect of it is equally challenging. This paper, based on fieldwork in Bolivia from 1975 to 1977, uses a case study approach to evaluate the adaptiveness of new settlement from the small farmers’ point of view along several parameters. Fieldwork yielded several
Traditional
215
types of data; reported here are the results of health interviews and a nutrition survey. Bolivia lies within the latitudinal tropics, but its extreme variations in altitude produce a great variety of environmental zones. Since prehistoric times, the human population has been concentrated on the altiplano, a high plateau which extends the length of the country, and in the high Andean r&es (see Fig. 1). However, about 60% of the national territory is made up of lowlands east of the Andes. The d~art~nt of Cochabamba, where fieldwork was conducted, contains some of each of the three major regions. Farmers from the Cochabamba valley, mainly from marginal agricultural areas, have been migrating to the nearby tropical Chapare lowlands for decades, but the trickle of migrants became a steady flow when an improved road linking the highlands and lowlands was completed in 1971. There are now about 33,000 people in the lowland portion of Cochabamba department, about 4% of the department’s population [2,3]. Over 80% of the lowland population is made up of spontaneous settlers and their families [33. MORBIDITY
AND MORTALITY
Historically, diseases have been an important deterrent to settlement of the tropical forests of lowland South America by highland peoples and Europeans [4,5]. One of the greatest disease threats, malaria, was eradicated from the Bolivian tropical settlement areas between 1950 and 1955 [6]. Several observers have predicted that new settlement will exacerbate disease problems by disrupting the tropical forest ecosystem and by bringing human populations into contact with little known disease agents [S]. A survey of health problems in one community in the Chapare, however, indicates that the poor health status of the settlers is due largely to intestinal parasitosis and associated anemia, problems which are common m%e tropics and well understood but persistent. In the community of spontaneous settlers where the data reported below were collected, there is no potable water supply; water is used directly from streams and even stagnant ponds. There are few latrines, and these are open shallow pits. The consequences of the poor sanitation practices brought from
216
CONNIE
WEIL
HIGHLAND, INTERMEDIATE, AND LOWLAND ZONES OF BOLIVIA 6’. $0 100.
0”
20
14’
40
IV
6’
16*
80
200 !o*
220 124
244 H4
l
lZ3
CAPITAL OF DEPARTMENT BELOW 5OOm - TROPICAL FOREST AND SAVANNA
@Zj 500 - 3000m l&J
- ANDEAN VALLES AND EASTERN SLOPES
ABOVE 3DOOm - ANDEAN ALTIPLANO AND CORDILLERAS
Fig. 1. The location of Cochabamba
department in relation to the altitudinal/ecological Bolivia.
the semi-arid highlands appear to be far more serious in the hot, humid lowlands. However, the health status of small farmers in the marginal areas of the Cochabamba valley also is poor. In evaluating health (or nutritional or economic) status relative to some optimum,, one must also ask whether or not the settlers’ situation represents an improvement compared with what their position would have been had they not migrated. In 1976 there were eight deaths in the research community, where the population grew from 225 to 254 during the year. This is a very small number of cases, but it does present a picture of the characteristics of mortality in the settlement zone (Table 1).
divisions of
Half of the deaths were of infants; these represented one-third of the babies born in 1976. In other words, of 12 babies born in the community in 1976, four had died by the end of the year. One adult died from pneumonia, and an adult and a child died from influenza, but all three had been weakened already by severe intestinal parasitosis and anemia. Only one of the eight deaths was caused by an “exotic” tropical disease. Since deaths are officially registered only sporadicaliy, there are no death rates or causes of death for the lowland settlement area as a whole. Further information on mortality is provided by the history of births and deaths of all children ever born to the research community’s property owners.
217
Morbidity, mortality and diet as indicators Table 1. Causes of death in the research community,
Cause of death
Under 1
Neonatal, unspecifiedt Amoebic dysentery Measles Influenza$ Pneumonia(j Penjgus uulgorisll
2 1 1 -
Total
1976 (N = 8)*
Age in years l-14 15 and over 1
1 1 1
2 I 1 2 1 1
1
3
8
-
4
Total
* Three other deaths occurred in the community during 15 months of fieldwork. An 8-year-old boy and an 18-month-old boy died after prolonged episodes of amoebic dysentery. A girl under 1 year of age died after a brief siege of amoebic dysentery. Only deaths occurring during the 1976 calendar year are presented in this table. t Two premature babies died within 48 hr of birth. 2 In May and June many residents contracted influenza, an unidentified viral infection. The symptoms included fever, headaches, back pains, and prostration. Most recovered without treatment. The two victims who died were a 28-year-old man and a 3-year-old girl, both of whom had amoebic dysentery and anemia before succumbing to the influenza. g The 55-year-old woman who died from pneumonia also had already been weakened by dysentery. 11 Penjigus vulgaris is a rare and poorly understood disease caused by a spirochete. The mode of transmission is unknown, treatment is symptomatic, and the prognosis for recovery is poor.
These data, based on structured interviews, are summarized in Table 2. Of the 202 children born alive, 18.8% died before age one and another 20.8% between the ages of one and four. Thus, nearly two-fifths died before age five. Numerous investigators suggest that the mortality rate among l- to Cyear-old children is a useful indicator of a population’s general health, particuiariy nutritional, status [7]. The infant mortality rate (deaths of children under 1 year of age) may be 10 times higher in a low-income country than in an industrialized country, but the l- to Cyear mortality rate may be 30-50 times higher. Thus, in industrialized countries the infant mortality rate is many times higher than the mortality rate among 1- to 4-year-old children. For example, in the U.S. it is about 25 times higher. Ages one through four include the transition to an adult diet. Since children of these ages have high nutrition needs for growth, they are particularly susceptible to the effects of an inadequate diet and parasitism. The worse the nutritional status of a population, the closer the infant mortality rate and the mortality rate among I- to 4-year-olds will be. For children of property owners in the research com-
munity, the very poor health situation is indicated by the fact that deaths in the one through four age group exceed infant deaths. Tables 3 and 4 help place the data on mortality of the landowners’ children in a broader context. Data for Bolivia as a whole are available from a national demographic survey conducted in 1976 [8]. It is estimated that 84.3% of Bolivian children survive to age one and 76.9% to age five (Table 3). This compares with 81.2% surviving to age one and 60.4% to age five in the research community. An interview survey carried out in 1975 by the Ministry of Public Health provides data for Cochabamba department [9]. These are reported here in a separate table because different age intervals are used. The survey found that 89.2% of the children in Cochabamba survive to age one and 83.3% to age six (Table 4). In rural Cochabambamore relevant to the present study since nearly all of the migrants come from rural highland Cochabamba-the figures are 86.4% and 78.6”/ respectively [lo]. The differences in survival rates to age one are not great enough to warrant firm conclusions. It is probably safe to say, however, that many more of the l- to 4-year-old children of property owners in
Table 2. Mortality of children born to property owners in the research community
Age in years
Number dying during age interval
Under 1 l-4 5-14 15 and over
38 42 4 2
Deaths during age interval as percentage of total live births 18.8% 20.8% 2.0% 1.O%
Number surviving at beginning of age interval 202 164 122 118
Deaths during age interval as percentage of number alive at beginning of interval 18.8% 25.6% 3.37, 1.7:;
218
CONNIEWHL Table 3. Survivorship at beginning of specified age intervals for Bolivia as a whole and for the research community Children surviving at beginning of age interval as percentage of total live births Age in years
Bolivia*
Under 1 l-4 5-14 15 and over
100% 84.3% 16.9% 74.4%
Research community 100% 81.2:; 60.4% 58.4%
* Source: Calculated from Somoza, op. cit. [S]. Table 4. Survivorship at beginning of specified age intervals for Cochabamba ment and for rural Cochabamba
depart-
Children surviving at beginning of age interval as percentage of total live births Age in years Under 1 l-5 6-10 I1 and over
Cochabamba
department
100% 89.2% 83.3% 82.5%
Rural Cochabamba 100% 86.4:; 78.6% ll.lO/,
Source: Bolivia. Ministerio de Salud Pliblica. op. cit. [9].
the research community have died than would be expected in Bolivia as a whole or in rural Cochabamba [ 111. DIET
Between late August and early November of 1976, a nutrition survey was carried out with 27 households in the research community [12]. Simple difference-ofmeans tests indicate that there is no significant difference between the mean values in age, years of residence in the community, years of residence in the lowland settlement zone, or wealth in manufactured goods for the residents included in the nutrition survey and tho& i&t. The major guide for methodology in the nutrition survey was Emma Reh’s Manual on Household Food Consumption Surveys [13]. Each household was visited once a day for seven consecutive days, and interviews were conducted with the person who ordinarily prepared food for the household. Information was solicited for the preceding 24 hr about how many meals had been eaten, who had been present for each one, and what food items had been included. It was possible to weigh quantities estimated by the informant for over SOo/,of the food items reported. Estimations for items that could not be weighed were based on a verbal description by the informant, then comparison with standards developed for the community. For each meal any leftovers or food fed to animals or thrown away was estimated. In addition, each member of the household was asked about food eaten between meals or away from home. Height/weight
indices
Of the 173 corisumers in the nutrition survey, 55 were age 17 or younger [14]. The heights and weights of these pre-adult individuals provide information
about their nutritional status. Tables 5 and 6 sum-. marize the results obtained when the height of each is expressed as a percentage of the reference or expected height for a person of his age and sex and when his weight is expressed as a percentage of the reference weight for a person of his height and sex. These percentages are based on World Health Organization standards developed for use with populations in Central America [15, 161. Waterlow and Rutishauser distinguish two results of inadequate nutrition. Stunting, commonly regarded as evidence of past undernutrition, refers to individuals with low height for age. Wasting, often regarded as evidence of current undernutrition, refers to individuals with low weight for height. Many individuals, of course, exhibit both types. Waterlow and Rutishauser feel that stunted children probably should be considered currently undernourished, even if their weight for height is normal [173. There is no fixed definition of “normal” vs stunting or wasting. Waterlow and Rutishauser suggest 90% of reference height for age and 80% of reference weight for height as dividing lines between mild and moderate stunting and wasting. They consider values below 85% of reference height for age and below 70% of reference weight for height as indicators of severe stunting and wasting [ 181. None of the children under age one in the nutrition survey exhibited stunting, i.e. low height for age (Table 5). However, nearly threequarters (14 of 20 children) of those in the l- to 4-year age group were moderately or severely stunted. Female toddlers were more likely than males to fall in these categories. In the 5- to 9-year age group, the only stunted children were females. There were a few older stunted children, but none in the “severely” stunted category. Only 2 of the 55 individuals age 17 or less exhibited wasting, i.e. low weight for height. These were
0
Total
-
-
-
5-9
IO-14 15-17
l I 5
--
2
0
-
-
-
-
0
2
2
-
-
80-90x
7@-8O%t
* Severe wasting, t moderate wasting.
Total
1-4
3
2
IO
2 -. 1 4 3
90-95x -
-
4
I 8
3 -
5
3
-
4 1 -
7
3 2 I I
-
2 I
-
1 -
85-90x$
75-80’/* ,0
lOO-105%
4 1 2
9s-100%
I 4
5
I
1
3
2
I
-
-
95-100x
90-95x
Females (N = 26)
-
-
6
I
4
I
90-lOOo/,
5
13
2
1
; -
,llO-120%
1
4
-j 5
1W110%
Males (N = 29)
-
z -
3
3
$2~130%
-
; -
I
60-700/6
1
_! -
70-8O%t
7 -
1
80-90%
7
_z. -
;
9O-100%
7
_! 1
1;
loo-110%
Females (N = 26)
4
I
_!
llO-120%
Table 6. Percentage of reference weight for height for individuals through age 17 (Figures represent number of individuals in each category)
60-70%.
Under I
Age in years
l
-
-
85-90x$
Males (N = 29)
8@-85%’
Severe stunting, $ moderate stunting.
-
5-9 lo-14 15-17
1-4
-_ -
75-f!o”/* $0
Under 1
Age in years
Table 5. Percentage of reference height for age for individuals through age 17 (Figures represent number of individuals in each category)
3 5
1
I
l20-130%
2
-
1 I
IOO-105%
220
CONNIE WEIL
l-month-old and 15month-old girls. The striking result reported in Table 6 is the high proportion of individuals whose weight is substantially higher than expected for their height. Youths between the ages of five and seventeen in the research community appear to have an adequate diet [19]. The height/weight data reported here, added to the mortality data, present a dismal picture of the health status of toddlers in the tropical settlement zone. No height/weight data for the highlands are available for comparison.
Calorie and protein consumption Data from the nutrition survey were converted to weights of edible portions of food items. Then food composition tables were used to calculate the calorie and protein content of foods actually consumed [20]. The adequacy of diet is evaluated by comparing actual consumption with estimated nutritional requirements [21]. The estimated requirements are based on sex and age and expressed as calories and grams of protein required per kilogram of body weight. Adjustments are made for pregnant and lactating women. For the research community as a whole, the per capita daily consumption is 2344 calories [22]. Based on the estimated requirements of individual consumers, actual consumption represents 98.4% of calorie requirements. These figures compare favorably with previous estimates for Bolivia of per capita daily consumption of 1870 calories, representing only 84% of estimated requ~ements [23]. In general, nutrition levels in BoIivia may have deteriorated sina: the time of these estimates due to the failure of agricultural production to keep pace with population growth and the inflation of food prices [24]. In the research community the per capita daily protein consumption was 57.4 grams. This is about the same as the average for Bolivia (57.5 grams) found in a national survey and substantially higher than the estimate for low altitudes in Bolivia (50.6 grams) from the same study [25]. Animal sources--meat, poultry, fish, eggs, and dairy products-accounted for 40.3% of the.protein consumed. This is reasonably’high by low-income country standards [26]. Since the unit of study in the nutrition survey is the household. data on consumption by individuals are not available. Even though calorie and protein consumption is generally adequate, all individuals do not necessarily get their share. Height-weight data and mortality figures indicate that l- to 4-year-old children. particularly females, in the research community do not [27J Economic aspects of diet So far the findings of the nutrition survey have been interpreted in terms of nutritional status, an important component of physical adaptation. Now the sources of calories and protein can be dealt with as indicators of economic adaptation. Peasants in the marginal farming areas from which the tropical settlers migrate raise potatoes, other tubers, and grains chiefly for home consumption, Surpluses are sold; but incomes are very low, and there is little money to purchase food. During numerous several-day visits to highland home villages of the
Purchasas 22.5%
(Therm add
OWll
/
SOURCES OF CALORIES up to 100.3% dua to rounding
SOURCES
OF
lrrors.1
PROTEIN
Fig. 2. Sources of calories and protein reported in the community nutrition survey in the Chapare tropical settlement zone. community’s residents. the only purchased foods observed being consumed were salt, sugar, and occasionally bread. Fuily 22.5% of the calories and 3 I.572 of the protein consumed during the nutrition survey were from purchased items (see Fig. 2). This represents a major difference in consumption patterns between small farmers in highland and lowland Cochabamba. It is one of several indicators of the much greater participation in the market ~onomy by settlers in the lowlands, Despite the importance of purchased food in the research community, the chief source is still home production, which accounts for 58.4% of the calories consumed and 48.2% of the protein. Food produced or, more accurately, procured by the farmers themselves includes both agricultural produce (which in turn inciudes pork, poultry, eggs, and milk as well as crops) and fish and game. Agricultural home produce accounted for 56.976 of total calorie consumption and 39.5% of total protein consumption. research
CONCLUSlONS
Although the settlers produce over half of their own food, they are far from self-sufficient. The large amount of food purchased indicates their high degree of involvement in the market economy. Since the reason peasants move to the lowlands is to raise their socioeconomic status, their increased incomes and buying power are legitimate measures of “success”.
Morbidity. mortality and diet as indicators Calorie consumption in the research community appears to meet the estimated requirements of community residents. Overall protein intake is considerably higher than estimated requirements, and a substantial portion of the protein consumed is from animal sources. Therefore, lowland settlement is “adap tive” in terms of nutritional status relative to that in the highland areas of out-migration. The mortality rate for l- to 4-year-old children, however, is apparently higher in the lowland settlement area than in the highlands. The price the settlers are paying for economic gains may very well include losing more of their toddlers. The situation outlined here for the Chapare tropical settlement zone may be analogous to that in early European cities, where economic incentives drew from the countryside the migrants necessary to replace the population ravaged by the diseases of urban congestion and insanitation [28]. Over time the health status of urban dwellers improved. With development of the same basic facilities-potable water supplies and sanitary waste disposal-the Chapare also may become a betted place to live. Acknowledgemenrs-This research was supported by a Fulbright-Hays fellowship. Jeanne and Paul Hicks assisted in the tabulation of and Xavier Pi-Sunyer in interpretation of the nutrition data. Figure 1 was prepared by Paul Heim and Figure 2 by Su-Chang Wang. Jim Weil served in multiple capacities.
REFERENCES
1. Nelson M.
The Developmenr in Latin America.
of Tropical
Lands:
Policy
Johns Hopkins University Press for Resources for the Future, Baltimore, 1973. 2. Bolivia. Servicio National de Formacibn de Mano de Obra. Estudio de Mano de Cobra en Cochabamba. p. III-l. FOMO. Cochabamba. 1977. 3. Galieguillos Fajardo A. La Colonizacidn en Bolioiu. Bolivia, Instituto National de Colonizaci6n. La Paz_ 1975. 4. Gade D. W. Inca settlement and endemic disease in the tropical forest. Unpublished paper presented at the Issues
74th Annual Geographers,
5. Goodland Green Hell
Meeting of the Association of American New Orleans, 1978. R. J. A. and Irwin H. S. Amazon Jungle: to Red Desert:’ pp. 48-55. Elsevier Scientific
Publishing, Amsterdam, 1975. 6. Reye U. Aspectos sociales de la colonizacibn del oriente Boliviano. Aporres 17, 56, 1970. Policy in 7. Aylward F. and Jul M. Prorein and Nutrition Low-Income Counrries pp. 21-23. Wiley, New York. 1975; DeMaeyer E. M. Clinical manifestations of malnutrition. In- Food, Man, and Society (Edited by Walcher D. N. er al.). D. 75 Plenum Press. New York, 1976: JellifTe D. B. ?ie Assessment of the Nurrirional Status of the Communir!, pp. 98-99. World Health Organization, Geneva. 1966; Wills V. G. and Waterlow J. C. The death rate in the age group 14 years as an index of malnutrition. J. rrop. Pediat. 3, 169-170. 1.958.
8. Somoza J. L. Encuesta demogrlfica national de Bolivia. Notas de Poblacidn 4. 31. 1976. 9. Bolivia, Ministerio de Salud Piblica. Cuadros Esradisricos: Encuesta Nurricional, pp. 41-46. Universidad Mayor de San Simon. Cochabamba, 1975. 10. A comparison of national data and those for Cocha-
221
bamba department indicate that both infant mortality and toddler mortality rates are lower in Cochabamba than in the country as a whole. This is probably accurate, since the standard of living even among peasants in rural Cochabamba is somewhat higher than in more isolated parts of the country. 11. Even this finding calls for caution, however, because of the small sample sizes and the fact that residents of the research community spend time in the highlands and cannot be defined strictly as lowland dwellers. Nevertheless, their primary residence during the child-bearing and raising years has been the tropical settlement zone, so the interpretation of mortality among their children as an indicator of relative adjustment to a new ,2. environment is probably sound. These represent nearly all full-time residents in the community. The survey was intended to include 28 households for whom detailed records of weekly income and expenses were being collected over a 12-month period; but one family chose not to participate in the nutrition survey. 13. Reh E. Manual on Household Food Consumption Surveys. United Nations, Food and Agriculture Organization. Rome, 1962. 14. There are more consumers than residents in the twentyseven households because visitors were counted as part-time members of the consumption group. 15. Organizacibn Mundial de la Salud, Instituto de Nutrici6n de Centro Amkrica y PanamB. Estcindares de Peso y Estatura. OMS, Guatemala, n.d. 16. Nutritionists do not agree whether there are significant genetic differences in height for different populations. If there are, standards for Central America may be more appropriate for use in South America than those based on the U.S. population. 17. Waterlow J. C. and I. H. E. Rutishauser. Malnutrition in man. In Ear/y Ma/nutrition and Mentul Developmenr (Edited by Cravioto J. er al.), p. 14. Almqvist and Wiksell, Uppsala 1974. 18. Ibid., pp. 20-21. 19. Waterlow and Rutishauser refer to children who have attained their expected weight for height but not their expected height for age as having “recovered” from malnutrition (p. 22). 20. Wu Leung W-T. Tab/u de Composicidn de Alimentos para Use en Amhrica Latinu. Organizacibn Mundial de la Salud, Instituto de Nutrici6n de Centro Am&ica y Panaml, Guatemala, 1961. 21. United Nations, Food and Agriculture Organization and World Health Organization. Energy und Protein Requiremenrs. FAO, Rome, 1973. 22. Maniac (Manihot esculenta) accounts for over twofifths of all calorie consumption reported in the nutrition survey. Manoic is nutritionally a poor food. but it has the advantages of easy cultivation and year-round harvestability. Rice is the second most important contributor to calorie intake. The two next most important contributors are produced in the highlands-alcoholic beverages and potatoes. 23. U.S. Department of Defense. Interdepartmental Committee on Nutrition for National Defense. Boliviun Nutrition Survey. USDD, Washington. 1964. 24. U.S. Agency for International Development Mission to Bolivia. Agricultural Development in Bolivia: A Sector Assessment (USAID. La Paz, 1974). p. 10: Wennergren E. B. and Whitaker M. D., The Srutus qf Bolirian Agriculture. pp. 54. 56. Praeger. New York. 1975. 25. U.S. Department of Defense. Op. cir. [23]. 26. As a matter of comparison. U.S. protein consumption averages 90 to 100 grams per capita per day. about two-thirds of which comes from animal sources. Sixty years ago animal sources provided about one-half of the protein in the U.S. diet. See Phipard E. F. Protein
222
CONNIE
and amino acids in diets. In Itnprouemenr of Protein Nutriture, p. 167. National Academy of -Sciences, Washington, 1974. 27. In many places toddlers and females in general do not
WEIL
receive food proportionate to their needs. See United Nations, op. iit. c213, p. 88. 28. For example, see McNeil1 W. H. Plagues and Peoples, p. 55. Anchor Press, Garden City, NY, 1976.