Content uploaded by Rocio Loria
Author content
All content in this area was uploaded by Rocio Loria on Jan 07, 2014
Content may be subject to copyright.
In the agroexport zone of Los Santos Zone in Costa
Rica, coffee is harvested by migrant labor. Most
migrants are from Panama and Nicaragua. We describe
migrants’ housing- and service-related health determi-
nants, with analyses of ethnicity, nationality and geog-
raphy. We used interviews, observation-based assess-
ments, and the Geographic Information System to
assess a population of 8,783 seasonal migrants and
1,099 temporary dwellings at a total of 520 farms
during 2004-2005. We identified determinants of poor
health including widespread deficiencies in the quality
of grower-provided dwellings, geographical isolation,
crowding, lack of radio and television, and deficient toi-
lets and cooking facilities. The indigenous and non-
Costa Ricans shared the poorest conditions. Reluctance
to use mainstream public health services was wide-
spread, especially among foreign and indigenous
migrants and the geographically isolated. Post-study,
researchers organized workshops for audiences includ-
ing workers, coffee producers, public officials and serv-
ice providers. Topics have included migration, preven-
tive health and hygiene, and child labor. This work was
successful in convincing Costa Rican social security
authorities to implement reforms that improve access
to and quality of health care for the migrants. Special
projects on ergonomics, psychosocial health hazards,
and water quality, as well as a literacy program, are
ongoing. Key words: migrants, indigenous populations,
seasonal work, living conditions, human rights.
INT J OCCUP ENVIRON HEALTH 2008;14:129–137
S
tudies have found a large variety of health risks
and hazards in migrant populations, including
those active in agriculture.
1–9
The hazards are
compounded by maltreatment of migrants on the basis
of ethnic and class differences and/or gender.
4,10–13
There is a lack of data on structural and contextual
determinants of migrant health in receiving countries/
communities. Various health-related selection and
mobility patterns of migrants, incomplete reporting of
illnesses and accidents, and long latency periods of
many health outcomes make direct health data hard to
interpret.
14–18
Underlying determinants of health are
usually more readily visible than their less easily-docu-
mented health effects. These determinants include a set
of overlapping conditions such as socio-economic and
class-related hazards,
11,13,19–24
cultural differences,
3,25–29
ethnocentrism and racism,
11,13,30
violence,
4,31,32
anti-
immigrant practices,
11,30,31
discrimination and abuse of
women,
4,13,21,31–34
lack of social protection,
25,39,34–36
pat-
terns of self-protection,
26
social and physical isolation,
9,26
unemployment, underemployment and precarious
work,
13,20,22
restricted opportunities for organization
and civic participation (that can be and have often
been overcome
37,38
), substandard housing and working
conditions,
1,5,9,28,36,39–41
food insecurity,
42
restricted
access to basic services,
28,34,35,39,43–45
and general struc-
tural marginalization and repression.
25,29,46,47
These
determinants operate at different levels and in differ-
ent combinations of conditions in migrants’ origin,
destination and on the journey; and can be studied
with quantitative and qualitative methods, including
ethnographic research that reaches deeply into con-
texts, social and economic mechanisms, power, atti-
tudes and behaviors.
4,11,48–50
In the agroexport zone of Los Santos in Southwest of
San José Province, Costa Rica, coffee is seasonally har-
vested by women and men who migrate, often with fami-
lies, from Panama, Nicaragua, and other locations in
Costa Rica (domestic migrants). The majority of Pana-
manian migrants are indigenous Ngöbe; Panamanian
and Costa Rican migrants include indigenous Ngöbe and
Bribri/Cabecar people. This survey is a product of the
first phase of a project entitled Community Empowerment in
the Rural Informal Sector: Work, Health and Socioeconomic
Intervention among Seasonal Coffee Harvesters in Los Santos,
Costa Rica, itself a component of the program Work and
Health in Central America (SALTRA; www.saltra.info). A
data-based description of working and living conditions
of migrant harvesters and their family members is essen-
tial to understanding migrant health challenges in Costa
Rica. This survey, which took place during the 2004–2005
harvesting season, identified a number of determinants
of poor health for the seasonal migrant coffee harvesters
during their temporary stay in Los Santos. This informa-
tion is crucial to the design of intervention and interac-
tion strategies intended to reduce health risks and
129
Received from: the Program on Work and Health in Central
America (SALTRA), Central American Institute for Studies on Toxic
Substances (IRET), Universidad Nacional (UNA) Heredia, Costa
Rica. Address correspondence to Dr. Timo Partanen, Central Amer-
ican Institute for Studies on Toxic Substances,Universidad Nacional,
Campus Omar Dengo, P.O. Box 86, CR-3000 Heredia, Costa Rica.
Disclosures: The authors report no conflicts of interest.
Determinants of Health in Seasonal Migrants:
Coffee Harvesters in Los Santos, Costa Rica
ROCÍO LORÍA BOLAÑOS, MSC, TIMO PARTANEN, PHD, MSC, MPH,
MILENA BERROCAL, MSC, BENJAMÍN ALVÁREZ, BH, LEONEL CÓRDOBA, BH
improve the living and working conditions of the itiner-
ant populations in the Zone.
MATERIALS AND METHODS
Los Santos is a mountainous zone in Costa Rica
between the capital San José (in Costa Rica’s Central
Valley) and the city of Quepos (on the Southwest
Pacific coast.) The zone consists of three cantons: Tar-
razú, Dota y León Cortes, with a total resident popula-
tion of 32,375 persons in 2000, and 12,380 hectares of
coffee cultivation in the same year.
51
This survey covered 520 coffee farms, 1,099 tempo-
rary grower-provided dwellings, and 8,783 seasonal
migrant workers (transborder and Costa Rican
migrants) and their family members during the coffee
harvesting season between December 20, 2004 and
February 25, 2005. The population was identified
through contacts with their coffee grower employers,
who are listed in the municipal records and the records
of coffee producers’ cooperatives. Investigators first
approached migrants at their dwellings during early
morning hours to introduce the project, obtain con-
sent, and agree on times for observation/interview
visits. Observation and interview visits were conducted
in off-work hours during late evening or at night.
Informed consents were obtained orally, as many sub-
jects had low literacy levels. Due to efforts to create con-
fidence and to agree on suitable visit times, there were
no refusals. Parents or other adults provided data on
children. The median time of interview and observa-
tion in a dwelling was 1 hour. All interviews were con-
ducted in Spanish, as indigenous families had enough
command of Spanish to handle the interviews.
Six investigators of Central American Institute for
Studies on Toxic Substances at Universidad Nacional
(IRET/UNA) completed the interviews of the
migrants and evaluations of their dwellings. The entire
transdisciplinary team represented expertise in
anthropology, geography, psychology, medicine, epi-
demiology, statistics, and informatics. Interviewers and
observers were the same persons, who were formerly
experienced in similar field projects in similar condi-
tions. Interview data were collected on basic demogra-
phy, migratory route, health and illnesses, health care,
and work. The interview form was an adaptation of a
standard form of the Social Security Authority of Costa
Rica (CCSS), consisting of structured, closed-ended
categorical questions. The form was adapted from the
original to better address migrant conditions. Partici-
pants were also asked if they currently experienced ill-
nesses that had been diagnosed by a physician. Their
verbal responses were classified according to a slightly
modified International Classification of Diseases IX
(ICD-9).
Housing conditions were assessed with a standard-
ized observation protocol that required categorical
classification of living and working conditions. For
example, the condition of the roof was assessed using
three exclusive categories:
Good: no cracks or fissures; adequate entry of air;
adequate, not old; material uncorroded; well attached
to the main structure. Regular: older but resistant to
precipitation and wind in spite of cracks and fissures;
keeps the indoor air environment acceptable. Poor:
loosening of attachments with entry of wind and pre-
cipitation; or considerable infiltration of water; higher
than recommended ventilation; usually old roof with
cracks and fissures.
Geographical Information System (GIS) software
was used to determine the slope of terrain for the loca-
tion of each temporary dwelling, as well as the distance
to the closest basic services (ranging from the very
basic small-scale small convenience store or to popula-
tion centers offering multiple services, whichever was
closest). Water quality was judged by observation, and
categorized according to criteria from the national
water quality standard. Housing data were dichoto-
mized for the analyses.
Two databases were constructed, one based on per-
sonal data and the other on dwelling data. Certain vari-
ables pertaining to the dwellings were also disaggre-
gated into the person database.
130 • Loría et al. www.ijoeh.com • INT J OCCUP ENVIRON HEALTH
Migrant Ngöbe woman harvesting coffee in Los
Santos, Costa Rica
Cross-tabulated means, medians, standard devia-
tions, ranges and percentages were calculated. To sim-
plify the description of the physical conditions of the
dwellings by way of variable reduction in the analyses of
associations between ethnicity, nationality and geo-
graphic location and the quality of the temporary hous-
ing, a factor analysis of dichotomized dwelling-related
determinants of health was performed. A sensitivity
analysis resulted in insignificant differences between
orthogonal (VARIMAX) and oblique (direct OBLIMIN)
rotation solutions; the VARIMAX solution was chosen.
Factor scores were calculated for comparisons between
ethnic and national groups. They were calculated for
each person, using the weights of the solution obtained
from the dwelling data. This process transformed each
new factor into a scale with mean = 0 and standard devi-
ation = 1, thus losing any normative aspect in the factor
scores in the description of the dwellings or migrants,
but allowing for internal comparisons between sub-
groups. Multiple linear regression and logistic models
were applied to ascertain associations of the dwelling-
related determinants with ethnicity, nationality and
geography. Due to the census-type data, inferential sta-
tistics such as confidence intervals were not calculated.
Only the regression models are reported with confi-
dence intervals to demonstrate the approximate mar-
gins of the coefficients.
RESULTS
The total number of migrants covered by the survey
was 8,783 (Table 1). Most of these migrated from
Panama (N=4762). Of Panamanians, the majority were
indigenous Ngöbe (4678). There were 1,882
Nicaraguans and 2,515 (394 indigenous) Costa Ricans.
To estimate total migrant population in the Zone
during the season, we secured population figures for
117 of the 161 farms not included in our study. Calcu-
lating a mean migrant population figure for the 117
farms with known populations, we estimated the
migrant population for the 44 farms for which we
lacked data. The total estimated migrant population
for the Los Santos Zone was 11,100. We were not able
to include all migrants in our study due to exception-
ally early maturation of coffee, which resulted in the
departure, therefore unavailability, of the migrants
before the expected end of harvest time.
The migrants were young (median age for women 17;
for men 21 years; ranges 0–76 and 0–89, respectively).
They traveled in groups of 4 persons on the average, usu-
ally in families (73%), often including children (35% of
the population was younger than 15 years of age). The
indigenous Ngöbe migrants in particular traveled in
families (94%). Most migrants (63%) were men. Educa-
tional level was relatively low: 30% of men and 40% of
women had not completed primary education. In the
Panamanian Ngöbe population, these percentages were
41% and 61%. The non-indigenous Costa Rican popula-
tion had achieved the highest educational level.
Reported lack of documentation (workers without
working permits in Costa Rica) was highest (49%)
among the Nicaraguans, while among the indigenous
Panamanian Ngöbe it was 5%.
Work
The farms typically employed ten workers, the biggest
over one hundred. On average, 7–10 persons stayed in
a temporary dwelling provided by the coffee pro-
ducer/farmer, depending on the district. Eighty-two
per cent (93% of those >4 years of age) of the popula-
tion covered by the survey was composed of harvesters
and other agricultural workers (Table 2). Workers were
paid per basket picked. Workdays spanned 8–9 hours
Monday through Friday, and 4 hours on Saturday, 2–3
months each year.
Dwellings
A large number of deficiencies and adverse conditions
were encountered in the living conditions of the
VOL 14/NO 2, APR/JUN 2008 • www.ijoeh.com Determinants of Health in Seasonal Migrants • 131
TABLE 1 Numbers of Seasonal Migrants Covered by
Survey, Los Santos, 2004-2005
Group N %
Panamanian Ngöbe 4,687 53.4
Costa Rican Ngöbe 351 4.0
Panamanian (80%) and
Costa Rican (20%) Bribri/Cabecar 54 0.6
Non-indigenous Panamanian 64 0.7
Non-indigenous Nicaraguan 1,882 21.4
Non-indigenous Costa Rican 1,721 19.6
Hondureña 2 0.0
Unknown 22 0.3
All 8,783 100.0
SALTRA team collecting data (geographic positioning,
interviews, observations) in coffee farms, Los Santos,
Costa Rica
migrants (Table 3). Fully a third of the dwellings were
classified as shacks or improvised structures or hovels,
including garages or shops that were furnished for the
migrants. Every fourth dwelling had insufficient sani-
tary facilities, with access to only a collective toilet or
the river for human waste disposal.
Clear structural deficiencies were observed. About
half of the dwellings had cracked and permanenetly
dirty floors, deficient ceilings/roofs, or poor lighting
and ventilation.
Cooking was done outside or in no fixed place in sev-
enteen per cent of the dwellings, affecting 20% of the
population. Cooking with firewood was very common
(80%), though in 68% of the dwellings where firewood
was used, gas or electricity were also used. For 708
dwellings, cooking was done inside with firewood, with
accompanying respiratory and burn hazards. Kerosene
and battery were reported as sources of energy for illu-
mination and home appliances by 2%. Refrigerators were
very rare (2%). Water quality was judged deficient or
poor for 38% of the dwellings. For 28% of the dwellings,
the source of water was rain, river or creek. Ten percent
of the dwellings lacked electricity. For almost 80% of the
dwellings, garbage remained either exposed (deposited
on the ground or rivers) or was burned.
One third of dwellings (and one third of migrants)
lacked radio, and two thirds lacked television. The non-
indigenous Costa Ricans had the best access, especially
to television (51%).
The number of persons in a dwelling varied between
1 and 35 (median 7). Number of persons in a bedroom
varied between 1 and 25 (median 3).
Distances and slopes of the dwellings were deter-
mined with reference to services and population cen-
ters. The distance by road and/or path to the nearest
general store, village shop or equivalent shop was 2 km
or more for 45% of the dwellings, and the slope was at
least 25 degrees for 58% of the dwellings. A total of
457 dwellings (41.9% of all dwellings) had both a min-
132 • Loría et al. www.ijoeh.com • INT J OCCUP ENVIRON HEALTH
TABLE 2 Jobs of Migrant Workers in Los Santos Zone, by Sex. Individual Data of Persons >4 Years of Age
Men Women All
______________________ ______________________ ______________________
Job N % N%N%
Harvester 4,664 94.3 2,384 89.6 7,048 92.6
Agricultural worker 55 1.1 16 0.6 71 0.9
Harvester / agricultural worker 8 0.2 — — 8 0.1
Harvester / domestic services/ 2 0.0 5 0.2 7 0.1
Sales 1 0.0 — — 1 0.0
Domestic services 9 0.2 70 2.6 79 1.0
Child care 17 0.3 17 0.6 34 0.4
Does not work 192 3.9 196 6.4 361 4.7
All 4,948 100.0 2,661 100.0 7,609 100.0
Temporary dwelling for migrant coffee harvesters, Los
Santos, Costa Rica
TABLE 3 Physical Conditions of the Temporary
Dwellings (N = 1,099)
N %
Shack, improvised, hovel 307 33.4
Kitchen: external or without fixed
location 183 16.6
Toilet: collective or river 288 26.2
Floor in poor condition 500 45.5
Ceiling in poor condition 531 48.3
Walls in poor condition 586 53.3
Poor illumination 478 43.5
Poor ventilation 685 62.3
Cooking with firewood or battery 879 79.9
No refrigerator 1,075 97.7
No radio 398 36.2
No television 769 69.9
No electricity 111 10.1
Source of water: rain, river or creek 306 27.8
Source of water: private 806 74.0
Deficient/poor water quality 415 38.0
Disposal of excreta: river or ground 11 1.0
Disposal of garbage: burned,
ground or river 870 79.1
Animals in unhealthy condition 205 18.6
> 3 persons per room 316 29.0
> 3 persons per bedroom 548 50.2
Distance to basic services
a
2,000+ m 218 45.3
Slope between dwelling and basic
services,
a
≥25 degrees 632 58.0
a
Real distance (meters) and slope (degrees) by path/road
between each dwelling and the nearest general store/village
shop.
imum distance of 2 km or more and the slope at least
25 degrees.
Ethnic, National and Geographic Inequalities in
Dwelling-related Determinants
Factor analysis of the physical housing conditions
(Table 4) revealed seven factors with eigenvalues >1.
These were interpreted as
1. Inferior quality of dwelling structures
2. Isolation
3. Crowding
4. Lack of toilet/adequate sanitary facilities
5. No radio, no television, no animals
6. Deficient basic installations
7. Deficient cooking facilities.
The factor scores, as composite indices of the hous-
ing conditions, were used in the analyses of associations
between ethnicity, nationality and geographic location
and the quality of the temporary housing, as described
on the next section.
The analysis of the factor scores (Table 5) revealed
that the indigenous lived in significantly unfavorable
housing conditions. This was seen in relation to all fac-
tors, with adjustment for nationality and geography.
The effect of ethnicity was strongest for crowding, but
was notable also for deficient cooking facilities and
other basic installations and for the overall poor type of
temporal dwelling. Foreign nationality was associated
with inferior quality of the structure of the temporary
dwelling and deficient cooking facilities, relative to
domestic Costa Rican migrants. Geographic differ-
ences had independent effects on all factors.
Health
At interview, 82% reported no illness. The most
common illnesses reported were respiratory (preva-
lence 4.2%), digestive (2.1%) and cardiovascular
(2.0%) illnesses, allergies (1.9%), and unspecific pains
and fever (1.9%). A total of 142 women were pregnant;
134 (94%) of them were not receiving prenatal medical
care. Of children, 85% were not receiving growth and
development check-ups, and 77% reported lack of vac-
cinations (81% in the Nicaraguans), with minor differ-
ences between girls and boys. The highest vaccination
rate was observed in non-indigenous Costa Rican chil-
dren; yet even among them 67% did not have their vac-
cinations up to date.
The clinic of the Social Security Institute of Costa
Rica (INS) was reported most frequently as a potential
source of health services. However, 35% responded
VOL 14/NO 2, APR/JUN 2008 • www.ijoeh.com Determinants of Health in Seasonal Migrants • 133
TABLE 4 Factor Pattern Solution
a
Factor Loading
________________________________________________________
Variable C 1234567
Shack, improvised, hovel .54 –.01 –.00 .08 .72
.03 .07 –.08
Kitchen: external or without fixed location .50 –.02 .13 –.07 .26 –.26 .31 .49
Toilet: collective or river .61 .04 .07 –.01 .76 .07 –.09 .09
Floor in poor condition .67 .80
.02 .05 –.01 –.04 .06 .01
Ceiling in poor condition .72 .83 .00 –.04 –.15 –.02 .01 .02
Walls in poor condition .72 .84
–.02 –.01 –.07 –.07 .01 .06
Poor illumination .51 .57 .13 .04 .33 .17 .18 .01
Poor ventilation .59 .73
.08 .04 .18 .12 .02 .08
Cooking with firewood .54 .14 .06 .08 –.01 –.07 .13 .62
No refrigerator .53 .12 –.06 .01 –.04 .31 –.13 .61
No radio .47 –.00 .05 .09 .20 .64 .01 .05
No television .43 .17 .06 –.02 .23 .54
.13 .14
No electricity .45 .16 .08 –.00 .09 .04 .61
.18
Source of water: rain, river or creek .61 .23 –.06 .04 .09 .10 .13 –.06
Source of water: private .22 –.03 .01 –.03 –.09 .05 .77
–.03
Deficient/poor water quality .42 –.12 .12 .05 –.16 .18 .11 .26
Disposal of excreta: river or ground .57 .07 –.08 .05 .14 .03 .09 –.07
Disposal of garbage: burned, ground or river .46 .13 –.04 .43
.00 .11 .39 –.24
Animals in unhealthy condition .48 .05 .01 .03 .24 –.63 –.01 .09
> 3 persons per room .93 .09 .95
–.02 .09 .06 .04 .05
> 3 persons per bedroom .93 .05 .96
.00 .01 .02 .05 .04
Distance to basic services
2
2,000+ m .76 –.03 –.00 .68 .05 .01 –.04 .09
Slope between dwelling and basic services,
b
≥ 25 degrees .84 .03 .01 .91 .02 .01 –.03 .02
Eigenvalue (rotation) 3.07 1.92 1.80 1.57 1.38 1.37 1.24
All variables coded 0/1. Contents of values 1 indicated. Loadings >0.4 underlined. Calculated of dwelling data.
C: Communality (regenerated from the data).
a
Extraction method: Principal components. Rotation: Orthogonal VARIMAX, with Kaiser normalization.
b
Real distance (meters) and slope (degrees) by path/road between each dwelling and the nearest general store/village shop.
“nowhere” to the question “Where would you go if you
got sick?” choosing instead self-care and self-medica-
tion. Only 4% of the population covered by the survey
received medical care at an INS clinic. The high pro-
portion of non-goers warrants a more detailed analysis,
considering a tendency toward self-care, care in fami-
lies, or among ethnocultural groups, reported in other
contexts.
27
We included gender in this analysis,
because, in contrast with the dwelling conditions,
which showed little difference between men and
women of the same nationality and ethnic group (as
they usually lived together), healthcare-seeking behav-
ior could vary by gender within national or ethic group.
We used a logistic model to estimate risk ratios (as
odds ratios) of the reported “Nowhere” response as a
function of the following determinants: nationality
(non-Costa Rican), ethnicity (indigenous), geography
(each of the three cantons; and distance and slope to
the services), sex, age, and having an illness. In the
models, each predictor was adjusted for the others.
The models showed that the non-Costa Ricans had
almost two times higher propensity not to seek outside
care when sick, compared with the Costa Ricans (Table
6). The indigenous respondents were also reluctant to
use community services provided by the Los Santos
communities. In particular, the Panamanian indige-
nous Bribri and Ngöbe had high percentages of
“Nowhere” responses (90.9 and 40.4 %, respectively).
The “Nowhere” response was also slightly elevated in
men, working-age migrants, migrants without illnesses,
and in migrants most isolated from services.
DISCUSSION
We found a number of widespread deficiencies in the
temporary living conditions of coffee harvester migrants
and their families in Los Santos. The major categories of
adverse health determinants were inferior quality of
dwelling structures, isolation, crowding, lack of radio
and television, substandard basic installations such as
toilet, and deficient cooking facilities. Some conditions
encountered were clearly violated Costa Rican Housing
Codes. The indigenous and non-Costa Ricans shared the
poorest conditions. Geographic differences within Los
Santos had independent effects on the determinants.
Health services were poorly accessed. Ninety-four per-
cent of the 142 pregnant women were not under med-
ical care. Reluctance to use Los Santos health services
was widespread (35%) and concentrated in foreign and
indigenous populations and those geographically iso-
lated.
Substandard housing of migrant labor may be
assumed a widespread phenomenon in general,
41
but it
is less frequently documented in a quantitative manner,
thus contributing to the invisibility of the conditions of
migrants. Epidemiologic studies of migrant workers
134 • Loría et al. www.ijoeh.com • INT J OCCUP ENVIRON HEALTH
TABLE 5 Linear Regression Models of Factor Scores as Functions of Ethnicity, Nationality, Geographic Location
Factor Indigenous Non CR Tarrazú Dota León Cortés R
1. Inferior quality of dwelling structures
B 0.12 0.34 0.43 REF 0.28 0.22
95% CI 0.07–0.16 0.28–0.39 0.36–0.49 0.21–0.35
2. Isolation
B 0.18 0.00 0.78 REF 1.09 0.33
95% CI 0.14–0.23 –0.05–0.05 0.72–0.84 1.02–1.15
3. Crowding
B 0.58 0.04 REF 0.17 0.02 0.30
95% CI 0.53–0.63 –0.01–0.09 0.10–0.23 -0.02–0.07
4. Poor dwelling: shack/hovel without toilet
B 0.30 0.04 REF 0.04 0.30 0.21
95% CI 0.26–0.35 –0.01–0.09 –0.02–0.10 0.26–0.34
5. No radio, no television, no animals
B 0.10 0.22 0.30 REF 0.56 0.20
95% CI 0.05–0.14 0.17–0.28 0.23–0.36 0.49–0.63
6. Deficient basic installations
B 0.31 -0.18 0.31 REF 0.06 0.19
95% CI 0.26–0.36 –0.24––0.13 0.25–0.33 –0.01–0.13
7. Deficient cooking facilities
B 0.38 0.20 0.09 0.40 REF 0.29
95% CI 0.34–0.43 0.15–0.25 0.04–0.13 0.33–0.47
CR: Costa Rican. B: Regression coefficient. CI: Confidence interval.
a
R: Multiple correlation coefficient. REF: Reference canton. All
determinants binary, coded (0,1).
a
Confidence intervals are presented exceptionally, although the data are census-type data. The CIs may be interpreted as indi-
cators of credible ranges of the ORs in a similar, freely defined “superpopulation.”
that focus on health and illness, must deal with formi-
dable empirical and analytic challenges posed by
health-based selection of migrants, by incomplete out-
come reporting and/or registration, and by latency
periods of illnesses, which usually preclude etiological
inference in the conventional cross-sectional preva-
lence data.
14–18
Data on determinants of health such as
housing are scanty. In addition, housing conditions
depend on the general geographic and climatic condi-
tions as well as on the housing level of the country and
locality and national legislation. The type and tempo-
rality of work and other economic factors further com-
plicate efforts to describe the health status of migrants.
It has been shown that Latino farmworker households
in North Carolina, U.S. have high rates of substandard
housing,
41
which generally parallels our findings. How-
ever, these studies are not necessarily generalizable to
migrants in all situations. On balance, comparisons
between data sets on housing quality of migrant farm-
workers from different countries or regions is less
informative than comparisons between the conditions
of migrants and the stable population resident in a par-
ticular community or nation.
The representativeness of our data may be evaluated
for cross-sectional coverage and stability over time.
First, all identified migrants in the farms that we visited
(8,783) were included. The number of persons missing
in the 117 farms that remained excluded for logistical
reasons was estimated at about 2,300. We did not see
any reason to suggest that the living conditions and
other health determinants differed based on study
enrollment and we believe our population reflected
the entire population.
As for the temporal aspect, the population that
arrives each year is different in number and composi-
tion. A core group arrives each year, with additional
migrants arriving according to factors including
demand and supply, economic condition of the farms,
pay, natural conditions, and national regulation con-
cerning migration. However, the material conditions in
Los Santos change more slowly: the temporary
dwellings and their environments stay as they are or
degrade over time until improvements are made. This
was a cross sectional study and did not reflect changes
in living condition or health service utilization over
time.
Undocumented migrants, who were numerous
during the undertaking of this survey, have poor social
protection. After the harvesting period of 2004–2005, a
new law came into force in Costa Rica that criminalized
the hiring of undocumented labor. Implementation of
this law has probably minimized the undocumented
labor migrant force in Los Santos, which may empower
the documented labor force to complain about defi-
ciencies. This change, along with other Costa Rican leg-
islation such as that concerning eradication of shack
(slum) dwellings, is expected to improve the living con-
ditions of the migrants on Los Santos and elsewhere in
Costa Rica.
The seemingly low prevalence rates of self-reported
illnesses have two likely causes. First, the population is
young as well as fit for migration and the ensuing hard-
VOL 14/NO 2, APR/JUN 2008 • www.ijoeh.com Determinants of Health in Seasonal Migrants • 135
TABLE 6 Percentage Responding “Nowhere” to Question “Where would you go in Los Santos if you fell ill?,”
by Selected Determinants. Results of Cross-tabulation and Multiple Logistic Model
% Responding
Determinant “Nowhere” RR
a
OR
b
95% CI
c
Foreign 39.5 1.80 2.09 1.84–2.38
Costa Rican 21.9
Indigenous 39.2 1.32 1.18 1.08–1.31
Not indigenous 29.8
Man 37.0 1.16 1.20 1.09–1.32
Woman 32.0
Distance to services
d
2,000+ m 37.3 1.13 1.15 1.01–1.30
Distance to services
d
up to 2,000 m 33.0
15–59 years of age 36.6 1.11 1.13 1.02–1.24
0–14 to 60–89 years of age 33.0
Does not have an illness 36.1 1.15 1.12 0.99–1.27
Has an illness 31.5
Slope of distance to services
d
25+ degrees 36.7 1.14 1.06 0.94–1.21
Slope of distance
d
to services up to 25 degrees 32.2
a
Risk ratio: (% with determinant)/(% without determinant)
b
Odds ratio: Risk ratio, slightly overestimated, adjusted for other determinants in the logistic model.
c
95% confidence interval de OR. These intervals are presented exceptionally, although the data are census-type data. The CIs may
be interpreted as indicators of credible ranges of the ORs in a similar, freely defined “superpopulation.”
d
Real distance and slope by path/road between each dwelling and the nearest general store/village shop.
ships (the healthy migrant effect). Second, symptoms
may have been underreported because of a cultural
construct that they led them to believe that they had
“no right to feel bad” because of the requirements of
everyday working conditions and migrant conditions in
general.
4,15,26,48,49
The high rate of persons who do not
seek healthcare when sick suggests a prominent role of
self-medication, self-care, and/or care in families or
among ethnocultural groups, as reported in other con-
texts.
27
Other possible explanations include lack of
resources, no ability to leave work, and/or lack of
knowledge of service location. A closer analysis of the
present data showed that rate of using municipal serv-
ices was lower in the indigenous, as well as in the non-
Costa Ricans. Physical isolation from services also
seemed to have played a role.
RESULTS LEAD TO ACTION
The disclosure of the results of the present study has
had a key role in exposing the conditions in Los
Santos.
52
This project has been conducted in accor-
dance with the explicit principles of the program Work
and Health in Central America (SALTRA; a program
with 15 projects in all seven Central American coun-
tries; www.saltra.info) which include inbuilt elements of
information sharing, awareness raising, participation,
and creation of coalitions and interaction/interven-
tion. Cross-cutting principles of the SALTRA program
include attention to gender, age, ethnicity, and infor-
mal workers; as well as empowerment, equity and trans-
parency. As part of the project we have completed over
20 local informational and educational workshops with
various audiences (workers, coffee producers, the resi-
dent population, community officers and service
providers) on migration, preventive health and
hygiene, work, child labor, and solutions to pervasive
problems. The project and its results have been pre-
sented in international congresses and publications.
52
We have successfully convinced the Costa Rican social
security authorities to implement reforms that improve
access to and quality of health care for the migrants. A
literacy program for 80 migrants is underway in four
farms. Community representatives, academics and gov-
ernment officials have developed government-spon-
sored inter-institutional commissions on the improve-
ment of migrant conditions and interregional
indigenous health promotion. Special projects on
ergonomics, psychosocial health hazards, and water
quality are ongoing. One doctoral and one master-level
thesis are being complemented on the project. Com-
munication channels have been established with the
International Labor Organization, International Orga-
nization of Migration, and the Pan-American Health
Organization.
We thank all the migrants and their families. José Carballo and
Darío Villegas participated in the data collection in the field.
Notes
1. Mobed K, Gold EB, Schenker MB. Occupational health prob-
lems among migrant and seasonal farm workers. West J Med.
1992;157:367-373.
2. García JG, Matheny Dresser KS, Zerr AD. Respiratory health of
Hispanic migrant farm workers in Indiana. Am J Ind Med.
1996;29:23-32.
3. Gadon M, Chierici RM, Rios P. Afro-American migrant farm-
workers: a culture in isolation. AIDS Care. 2001;13:789-801.
4. Loría Bolaños R. De Nicaragua a Costa Rica y a Nicaragua. La
ruta crítica de las mujeres migrantes nicaragüenses: una mirada
desde la zona norte fronteriza [From Nicaragua to Costa Rican
and back to Nicaragua: The key route of Nicaraguan migrant
women: a view from the Zona Norte]. San José, Costa Rica:
Centro de Estudios y Publicaciones Alforja; 2002. Spanish.
5. Magaña CG, Hovey JD. Psychosocial stressors associated with
Mexican migrant farmworkers in the Midwest United States. J
Immigr Health. 2003;5:75-86.
6. Lukes SM, Simon B. Dental decay in southern Illinois migrant
and seasonal farmworkers: An analysis of clinical data. J Rural
Health. 2005;21:254-258.
7. Rabinowitz PM, Sircar KD, Tarabar S, Galusha D, Slade MD.
Hearing loss in migrant agricultural workers. J Agromedicine.
2005;10:9-17.
8. Cameron L, Lalich N, Bauer S, Booker V, Bogue HO, Samuels
S, Steege AL. Occupational health survey of farm workers by
camp health aides. J Agric Saf Healt.h 2006;12:139-153.
9. Hiott AE, Grzywacz JG, Davis SW, Quandt SA, Arcury TA.
Migrant farmworker stress: mental health implications. J Rural
Health. 2008;24:32-39.
10. Montenegro RA, Stephens C. Indigenous health in Latin Amer-
ica and the Caribbean. Lancet. 2005;367:1859-1869.
11. Holmes SM. An ethnographic study of the social context of
migrant health in the United States. PLoS Med. 2006;3:1776-
1793.
12. Ahonen EQ, Benavides FG, Benach J. Immigrant populations,
work and health – a systematic literature survey. Scand J Work
Environ Health. 2007;33:96-104.
13. Llácer A, Zunzunegui MV, del Amo J, Mazzarasa L, Bolumar F.
The contribution of a gender perspective to the understanding
of migrants’ health. J Epidemiol Community Health 2007;61,
Suppl. 2:ii4-10.
14. Abraído-Lanza AF, Dohrenwend BP, Ng-Mak DS, Turner JB.
The Latino mortality paradox: a test of the “salmon bias” and
healthy migrant hypotheses. Am J Public Health. 1999;89:1543-
1548.
136 • Loría et al. www.ijoeh.com • INT J OCCUP ENVIRON HEALTH
Child labor. Panamanian Ngöbe migrant in Los Santos,
Costa Rica.
15. Vulpiani P, Comelles JM, Van Dongen E. Health for all, all for
health. European experiencies on health care for migrants.
Perugia: Cidis Alisei; 2000.
16. Romani O. La salud de los inmigrantes y la sociedad. Una visión
desde la antropología [Immigrant health and society: A view
from Anthropology]. Formación médica continuada en aten-
ción primaria. 2002; 9:498-504. Spanish.
17. Franzini L, Ribble JC, Keddie AM. Understanding the Hispanic
paradox. Ethn Dis. 2001;496-518.
18. Fennelly K. The “healthy migrant” effect. Minn Med. 2007;
90:51-53.
19. Weio M, Valdez RA, Mitchell BD, Haffner SM, Stern MP, Hazuda
HP. Migration status, socioeconomic status, and mortality rates
in Mexican Americans and non-Hispanic whites: the San Anto-
nio Heart Study. Ann Epidemiol. 1996;6:307-313.
20. Beck U. La sociedad del riesgo global [World Risk Society].
Madrid: Siglo XXI; 2001. Spanish.
21. Östlin, P. Gender inequalities in health: the significance of
work. In: Wamala S, Lynch J. Gender and social inequities in
health. Lund: Studentlitteratur; 2002. p. 43-60.
22. Bauman Z. Wasted lives. Oxford: Blackwell; 2004
23. Díaz-Salazar R. Trabajadores precarios [Precarious Workers]. El
proletariado del siglo XXI. Madrid: Ediciones HOAC; 2004.
Spanish.
24. Franzini L, Fernández-Esquer ME. Socioeconomic, cultural and
personal influences on health outcomes in low income Mexi-
can-origin individuals in Texas. Soc Sci Med. 2004;59:1629-1646.
25. Blanco MC. El inmigrante como sujeto marginado: Claves inter-
pretativas [The immigrant as marginalized subject: Key inter-
pretations]. In: Alvite JP. Racismo, antirracismo e inmigración
[Racism, antiracism, and immigration]. Donosita: Gakoa Libu-
ruak; 1995. p 69-88. Spanish
26. Atxotegui R. Los duelos de la migración: una aproximación psi-
copatológica y psicosocial [The sorrows of migration: A psy-
chopathological and psychosocial approximation]. In:
Perdiguero E, Comelles JM. Medicina y cultura: Estudios entre
la antropología y la medicina [Medicine and Culture: Studies in
anthropology and medicine]. Barcelona: Edicions Bellaterra;
2000. p. 83-100. Spanish.
27. Haro JA. Cuidados profanos: una dimensión ambígua en la
atención de la salud [Profane citizens: An ambiguous dimen-
sion to healthcare]. In: Perdiguero E, Comelles JM. [Medicine
and Culture: Studies in anthropology and medicine].
Barcelona: Edicions Bellaterra; 2000. pp. 101-161. Spanish.
28. Hansen E, Donohoe M. Health issues of migrant and seasonal
farmworkers. J Health Care Poor Underserved. 2003;14:153-
164.
29. Grondin D. Well-managed migrants´ health benefits all. Bull
World Health Organ. 2004;82:561.
30. Sandoval C. Otros amenazantes. Los nicaragüenses y la forma-
ción de identidades nacionales en Costa Rica [Threatened
others: Nicaraguans and national identity formation in Costa
Rica]. San Pedro, Costa Rica: Ediciones Universidad de Costa
Rica; 2003.
31. Gregorio C. Migración femenina: su impacto en las relaciones
de género [Female migration: Its impact on gender relations].
Madrid: Nancea; 1998 Spanish.
32. Loría-Bolaños, R. Migración femenina y violencia de género:
panorama actual y escenarios de mujeres centroamericanas
[Women’s migration and gender violence: Current panorama
and scenarios for Central American women]. In Piqueras,
Andrés. Mediterránia Migrant II. Cooperació i solidaritat, pro-
jectes, 9. Castelló de la Plana: 255-270. Spanish/Catalan.
33. Gargallo F. Mujeres migrantes: pobreza real y discriminación de
género [Migrant women: Real poverty and gender discrimina-
tion]. Cuadernos Americanos. 2005;3:11-24. Spanish
34. Weathers A, Minkowitz C, O´Campo P, Diener-West M. Health
services use by children of migratory agricultural workers:
Exploring the role of need for care. Pediatrics. 2003;111:956-
963.
35. Ahmad F, Shik A, Vanza R, Cheung AM, George U, Stewart DE.
Voices of South Asian Women: Immigration and mental health.
Women Health. 2004;40:113-130.
36. Kim-Godwin YS, Bechtel GA. Stress among migrant and sea-
sonal farmworkers in rural southeast North Carolina. J Rural
Health. 2004;20:271-278.
37. Fox J. Mexican migrant civic participation in the United States.
New York,: Social Science Research Council, 2007. Available
from: http://borderbattles.ssrc.org/Fox).
38. Dyck I, Dossa P. Place, health and home: gender and migration
in the constitution of healthy space. Health Place. 2007;13:691-
701.
39. Slesinger DP. Health status and needs of migrant farm workers
in the United States. J Rural Health. 1992;8:227-234.
40. Arcury TA, Quandt SA, Barr DB, Hoppin JA, McCauley L,
Grzywacz JG, Robson MG. Farmworker exposure to pesticides:
methodologic issues for the collection of comparable data. Env-
iron Health Perspect 2006;114:923-928.
41. Gentry AL, Grzywacz JG, Quandt SA, Davis SW, Arcury TA.
Housing quality among North Carolina farmworker families. J
Agric Saf Health. 2007;13:323-337.
42. Weigel MM, Armijos RX, Hall YP, Ramirez Y, Orozco R. The
household food insecurity and health outcomes of U.S.-Mexico
border migrant and seasonal farmworkers. J Immigr Minor
Health. 2007;9:157-169.
43. Coughlin SS, Wilson KM. Breast and cervical cancer screening
among migrant and seasonal farmworkers: A review. Cancer
Detect Prev. 2002;26:203-209.
44. Murray SB, Skull SA. Hurdles to health: immigrant and refugee
health care in Australia. Aust Health Rev. 2005;29:25-29.
45. Arcury TA, Quandt SA. Delivery of health services to migrant
and seasonal farmworkers. Annu Rev Public Health. 2007;
28:345-363.
46. Bassin D. La patetización del mundo: ensayo de antropología
política del sufrimiento [The patheticization of the world: An
essay in the politcal anthropology of suffering]. In: Viveros M,
Garay G. Cuerpos, diferencias y desigualdades. Bogotá: Edición
Universidad Nacional de Colombia; 1999. pp. 31-41.
47. Tam CC. Migration and health: fact, fiction, art, politics. Emerg
Themes Epidemiol. 2006;3:3-15.
48. Martínez U. El Ejido: Discriminación, exclusión social y racismo
[The ejido: Discrimination, social exclusión and racism].
Madrid: Catarata; 2001. Spanish.
49. Lara Flores SM. Violencia y contrapoder: una ventana al mundo
de las mujeres indígenas migrantes, en México [Violence and
resistance: A window onto the world of indigenous migrant
women in Mexico]. Rev Estud Fem. 2003;11:381-397. Spanish.
50. Esteban ML. El estudio de la salud y el género: Las ventajes de
un enfoque antropológico y feminista [The study of health and
gender: The advantages of a anthropological and feminist
focus]. Salud Colectiva. 2006;2:9-20.
51. Instituto del Café de Costa Rica (ICAFE). Informe sobre la
actividad cafetalera de Costa Rica [Report on coffee-growing
activity in Costa Rica]. Congreso Nacional Cafetalero Ordi-
nario. San José, Costa Rica: ICAFE; 2004. Available at http://
www.icafe.go.cr/icafe/downloads/acc2004.pdf. Spanish
52. Loría Bolaños R. Empoderamiento del sector informal rural
[Empowerment of the rural informal sector]. Salud & Trabajo.
2007;4:6-7. Spanish.
VOL 14/NO 2, APR/JUN 2008 • www.ijoeh.com Determinants of Health in Seasonal Migrants • 137
