The role of household pests in the epidemiological transition of allergy


Material Information

The role of household pests in the epidemiological transition of allergy modernization of the domestic environment in Barbados
Physical Description:
ixx, 422 leaves : ill. ; 29 cm.
Barnes, Kathleen Carole, 1960-
Publication Date:


Subjects / Keywords:
Allergy -- Barbados   ( lcsh )
Household pests -- Barbados   ( lcsh )
Allergy -- Epidemiology -- Barbados   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1992.
Includes bibliographical references (leaves 398-421).
Statement of Responsibility:
by Kathleen Carole Barnes.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001867924
notis - AJU2440
oclc - 28997607
System ID:

Full Text







Copyright 1992


Kathleen Carole Barnes

This thesis is dedicated to the children of Barbados.


The author would like to thank the following individuals

and organizations in Barbados for assistance in administering

this project: the Honorable Branford Taitt (Minister of

Health), Dr. Beverly Miller (Chief Medical Officer), Dr. E.

Ferdinand (Acting Chief Medical Officer during 1991), and all

of the medical staff at the polyclinics in which the study

took place. The author is grateful to Drs. Timothy Roach,

Malcolm Howitt, David Corbin, and Raana Naidu for their

assistance in selecting the children for the study. Thanks

also go to staff members at the Entomology Section of the

Ministry of Agriculture and the Leptospira Laboratory for

their many means of assistance throughout the year.

The author is indebted to Drs. C.O.R. Everard (Leptospira

Laboratory), J.C. Hudson (Carib-Agro Industries, Ltd.),

Michael Nathan (PAHO), and Jeffrey Jones (Ministry of

Agriculture) for advice, direction, and support.

The author is grateful for a Graduate Research

Assistantship and supplies from USDA/ARS and to Dow-Elanco for

additional financial support. Thanks are extended to Mrs.

Margaret Haile and Mr. Curtis Guinyard for technical support.

Identification of arthropods was graciously provided by Mr.

Lloyd Davis (USDA/ARS) in Gainesville, FL, and Dr. Enrique


Fernandez-Caldas and Mr. Walter Trudeau (USF) in Tampa, FL.

Thanks go to Dr. Rick Helm (Arkansas Children's Hospital) for

methodological advice, Ms. Chris Anderson (FDA) and her staff

for conducting the serum assays, and to Mr. Kenneth Pope

(Baskerville & Sons, Richmond, VA) for the blueprints and

sketches. Additional gratitude is extended to Mr. Richard

Wadley (S.C. Johnson) and Mr. Walter Short (Geddes Grant

(Barbados) Ltd) for logistical assistance. Statistical

assistance was graciously provided by Ms. Cindy Hewitt & Dr.

David Nickerson (Orlando, FL), Mr. Victor Chew (Univ. of FL,

IFAS), and Ms. J. White (USDA, Gainesville, FL).

The author extends her appreciation to a warm and

supportive doctoral committee, including Drs. George

Armelagos, Richard Brenner, Leslie Lieberman, Linda Wolfe, and

Richard Patterson, and especially to her supervisory chair,

Dr. Gerald Murray, for his confidence, insight, and direction.

Very special thanks go to Dr. Richard Brenner, for his dynamic

and steadfast mentorship, friendship and multi-faceted


Lastly, the author offers her deepest gratitude to her

friend, Mr. David E. Milne, without whose technical and

emotional support this work could never have been completed,

and to her parents, Robert and Carole Barnes, sister,

Christine, and brother-in-law, Kenneth Pope, whose endearing

support and belief in the author saw this project to its end.





ABSTRACT .. . ..

. xii

. xvi

S xviii



Introduction: Rationale for the Study .
The Relationship Between Modernization
and Allergy .. ..........
The Role of Household Pests .
Objectives of the Study . .
Defining the Epidemiological Transition .
Disease Ecology and the Evolution
of Human Allergy . ..
Disease in Pre-State Society .
Disease in State Society .
Application of the Disease Ecology
Model: Allergy . .
The Influence of Disease on Biological
Adaptation: Parasitic Infestations
and IgE .. . .
The Influence of Disease on Cultural
Adaptation: The Gradual Decrease
in Exposure to Parasites .
Influencing Biological and Sociocultural
Factors . . .
Genetic, MHC-related factor .
Genetic, non-MHC-related factor .
Familial factors . .
Ethnic factors ...
Behavioral and nutritional factors
Miscellaneous behavioral factors .
Population behavior and the
macroenvironment . ...

. 1

. 1

. 1
S. 3
S. 6
. 8

. .

Population behavior and the
microenvironment . 40
Conclusions and the Allergy Disease Model 40
Notes. ... . .. 42


The Construct of Allergy and Modernization:
A Review of the Objectives . .. 44
The Research Site . . 46
The Rising Incidence of Asthma
in Barbados . .. 47
Infrastructure: The Developed Nature
of a Developing Country. ... 55
The Health Care Delivery System .. 57
Sampling and Methodology . 64
The Survey Schedule . 64
Topographical considerations 64
Selected entomological and
ethnographic research sites 70
Participant Selection and Resources .. 71
Clinical Testing and the Interview Schedule 74
Sub-Setting the Sample Population .. 78
Identifying and Classifying the
Independent Variables . 79
Designing a Socioeconomic Indicator 80
Income . 81
Occupation ......... 82
Amenities ...... 83
Monthly expenses . .. 86
"House-related" variables 87
Creating the wealth index .. .. 88
Demographic Characteristics of
the Sample Population .. ... 92
Spatial Distribution . 92
Personal and Socioeconomic Characteristics 96
Analysis and Presentation . 98
Summary ................ .... 99
Notes . . 99


A Historical Review of Land Distribution .. 101
The Birth of Barbadian Society:
The Pre-Emancipation Period .. 102
Land Allocation and Stratification of
the Planter Class . .. 103


Indentured Servitude and a White Lower Class 110
African Slaves: An Alternative
Labor Source . 113
Acculturation of the Classes 115
Emancipation and the Landless Proletariat 117
Steps Towards Reform . 121
The Contemporary Spatial Distribution of
Land and People . 123
The Impact of Twentieth-Century Reforms 123
Barbadian Economy and Its Role in the
Urban/Rural Sector . 125
The Over-Development of the Urban Sector 129
Inter-Migration and Changes in the
Spatial Distribution of Land and People 130
Demographic Decentralization and
the Development of Suburbia .. 131
The Continuation of Tenantry Residency 140
A Closer Look: Evolution of the
Barbadian Home 143
Traditional House Forms .. 144
Housing in the pre-emancipation era 144
Emancipation and the chattel house 146
Other house forms and modifications
of the chattel house . 150
Trends in Contemporary Housing . 156
The Role of Ownership . 156
A decline in relocation 160
Upgrading the chattel and creating
new forms . 161
The transition from wood to masonry 168
The problem of construction in the
informal sector . 176
Elaboration of the indoor environment 181
Installation of indoor utilities 182
The Role of the Family System ...... 185
The impact of costs and tenure 185
Trends in household density 187
Housing Trends and Health Implications 188
Moisture . .. 188
Ventilation ...... 197
Summary ................... 199
Notes. . . 201


Spatial Distribution of Land
and Sociodemographics. . .. 209
The "Ideal" Place of Residency. .. 216
Implications for Trends in Housing 220


Housing Patterns Among the Asthma Study
Population . .. .. 222
Ownership of Dwellings. . 222
Construction of Dwellings . 225
Structural Features and Demographics 230
Aspirations of the Study Population:
The "Ideal" House . 231
Discussion and Summary . .. 235
Notes . . 238


Introduction ...... ...... 242
Infamous Household Pests in Barbados .. ... 244
Pests Related to Disease .. .... 244
Mosquitoes .. ... 245
Other diptera ..... ... 247
Cockroaches ......... 247
Millipedes .. .. .... 248
Centipedes .. 249
House dust mites .. 249
Rodents ....... 251
Pests not Related to Disease .. 253
The Pest Collection: An Inventory
of Household Pests . 255
Materials and Methods 255
Light/CO, trap .. 256
Flea trap ......... 257
Fly strips ........ .... 257
Cockroach traps .......... .257
Live trap . 258
Mite sampling ........ 258
Results: Collective Taxonomy
of Barbadian Household Pests .. 260
Rodents and other mammals ... 261
Orthoptera .... ...... 263
Coleoptera ....... 264
Amphibians/reptiles .. 264
Diptera . .... 265
Hymenoptera .. . 266
Isoptera o . 266
Lepidoptera ............ 266
Isopods .......... 267
House dust mites .. .. ..... 267
Discussion of the Pest Collection Results 271
Food and/or garbage pests .. 271
Pests dependent on human hosts 272
Moisture-related pests 273

Stored food pests . .
House dust mites . .
The Ethnographic Survey: Categorization,
Classification, and Responses
to Household Pests . .
The Ethnographic Taxonomy .
Reported Temporal Distributions .
Interrelationships with Sociodemographic
and Structural Variables .
Rodents . .
Cockroaches . .
Mosquitoes . .
Sand flies . .
Others . .
Classifying the Pests:
What's Good and Bad? .. .
Which of these are "disgusting"? .
Which of these are "harmful"? .
Which of these pests would
you kill? .
Which of these are "beneficial"? .
Are these pests indoor, outdoor,
or accidental? . .
Behavior Directed at Controlling
Household Pests . .
National Control . .
Commercial Control . .
Control in the Private Sector .
The Effect of National Campaigns Directed
at Community Control . .
Summary . .
Notes . .


Biomedical Testing . .
Biomedical Materials .
Subjects . .
Extracts . .
Biomedical Methods .
Interviews . .
Skin Test Results . .
Individual Extracts .

House dust mites .
Crawling insects .
Flying insects .
Shellfish ..
Rodents ..

. 341

S 344
S 344
S 347










. .
. .
. .
. .
* .
* .
. .
. .

Cross-Reactivity . .
Discussion . ..
Symptomatology . .
Individual Allergens . .
Cross-Reactivity . .
Summary . . .
Notes . . .


Asthma as a Disease of Modernization:
Concurrence with the Theoretical Construct .
Review of the Findings . .
Modelling the Variables . .
Implications for Community- and Household-Level
Management of Pests . .
Implications for Patient Education .

Financial constraints . .
Structural constraints .
Environmental constraints .
Ideological constraints ..
Implications for Further Research .





S. 389
S. 390
S. 392
S. 393
. 394






Table 2-1 GNP per capital (1980), in U.S. dollars in
the Eastern Caribbean . 56

Table 2-2 Health indicators for the Eastern
Caribbean . . 60

Table 2-3 Profile of education in the Eastern
Caribbean . .. 61

Table 2-4 Polyclinic/outstation attendances, 1990 63

Table 2-5 Frequency distribution of amenities by
household (N=177) . .. 85

Table 2-6 Distribution of total monthly expenses in
the household sample (N=177). ... 87

Table 2-7 Distribution of residences by demographic
region (N=177) . . 92

Table 2-8 Spatial Distribution of the Total Asthma
Study Population (N=335). . 95

Table 2-9 Frequency and Percent of Types of Unions
per Informant (N=177) . .. 97

Table 3-1 Percentage of gross domestic product (GDP)
by sector. ... . 129

Table 3-2 Percentage distribution of the Barbadian
population by socio-geographic zone, from the
Government Statistical Department Census for
1970 and 1980 . . 137

Table 3-3 Spatial distribution of tenantries by
parish, 1990. . .. 140

Table 3-4 Tenants living in National Housing
Corporation units by parish, 1991 ... .159

Table 3-5 Dwelling tenure (owned, government rental,
or other) by parish, 1980 . .. 163


Table 3-6 Number of houses relocated by parish,
1981-1989. . .. .. 164

Table 3-7 Number and percentage of dwellings by
type of material, for each parish, 1980 169

Table 3-8 Applications approved by the Town & Country
Development Planning Office for renovation of
dwelling units, by parish, 1980 .. 170

Table 3-9 Number of new and renovated dwellings by
house type, approved by the Environmental
Division, Engineering Ministry of Housing &
Lands, 1982 1989 . 173

Table 3-10 Number of persons per household by
parish, for the years 1946, 1970, 1980,
and 1990 . . 189

Table 3-11 Percent of households calculated for the
number of persons by the number of rooms 191

Table 3-12 Sources of moisture in the home .. 192

Table 4-1 Land ownership patterns according to
demographic region (N=177) . ... 211

Table 4-2 Frequency and percentage of households
owning their residential lot, by parish 213

Table 4-3 Distance from the nearest neighbor
(in yards) according to demographic region
(N=177) . . 214

Table 4-4 Analysis of residential preference by
parish among households . .. 219

Table 4-5 Overall rank of parishes as "the ideal
place to live" according to the total informant
population (N=166) . .. 220

Table 4-6 Home ownership patterns according to
demographic region (N=177) . .. 224

Table 4-7 Frequency and percentage of informants
owning their dwelling, by parish .. 226

Table 4-8 Comparison of housing quality (concrete
or mixed structure) by wealth score, home
ownership, and land ownership .. 227


Table 4-9 Structural features of the dwelling by
the type of material the dwelling is constructed
and for the total number of dwellings, for the
asthma study households . .

Table 5-1 Density of house dust mite species D.
pteronyssinus and D. farinae in 17 Barbadian
homes . . ..

Table 5-2 Reported pests in Barbadian homes by
informants in the asthma study. .

Table 5-3 Ethnographic taxonomy of 33 pests
(frequency and percent of responses, N=65).

Table 5-4 Frequency and percentage of responses to
the question: "Which pest do you have the
biggest problem with?" . .

Table 5-5 Frequency and percent of specimen
categorization regarding when informant sees the
animal around the house (N=65). .

Table 5-6 Responses (frequency distribution and
percentages) to the question: "Which of these
pests do you consider to be 'disgusting'?" .

Table 5-7 Responses (frequency distribution and
percentages) to the question: "Which of these
pests can cause physical harm to people, and
how?" (N=65) . . .

Table 5-8 Responses (frequency distribution and
percentages) to the question: "Which of these
pests would you kill if you saw it?" (N=65)

S 317

S 318


Table 5-9 Correlations between pests that were
classified as "disgusting," and those that the
informant "would kill" (Chi-square) .

Table 5-10 Responses (frequency distribution and
percentages) to the question: "Which of these
pests are beneficial, or good to have around?",
and "Why?" . . .

Table 5-11 Pests for which over-the-counter
insecticides are used, by frequency (percent) of
the informant population (N=65) .











* .

Table 6-1 Relationship between allergies to house
dust mites, D. pteronvssinus, and D. farinae,
and structural and demographic variables. 348

Table 6-2 Total number and proportions of responses
to specific extracts. ......... 352

Table 6-3 Frequency distribution for total numbers
of different pests reacted to in the battery of
skin testing. . . 353

Table 6-4 Frequency distribution of stimulants
responsible for an asthmatic attack (N=168).. 360

Table 6-5 Time period when child is reportedly most
likely to experience an asthmatic attack 361


Figure 1-1 The Allergy Disease Model in the
Ecological Perspective .. . 43

Figure 2-1 Asthmatic attendances to the Accident &
Emergency Department, Queen Elizabeth Hospital,
in thousands, for 1980 1990. . 49

Figure 2-2 Number of asthmatics attending the
Accident & Emergency Department, Queen Elizabeth
Hospital, for the month of October, 1974 1990 50

Figure 2-3 Monthly Asthmatic Attendances to the
Accident & Emergency Department Compared to Monthly
Rainfall, 1990 . . 51

Figure 2-4 Monthly Asthmatic Attendances to the
Accident & Emergency Department Compared to Monthly
Rainfall, 1991 . . 52

Figure 2-5 Distribution of outpatient health care
facilities . . 62

Figure 2-6 Geophysical Subdivisions of Barbados 66

Figure 2-7 Annual total rainfall by region, in
inches, based on 1887-1986 data .. 69

Figure 2-8 Location of selected communities for the
entomological and ethnographic surveys 75

Figure 2-9 Resource centers for asthma study sample 76

Figure 2-10 Non-linear regression analysis
demonstrating the relationship between the
variables "monthly house rent/mortgage" and
"amenities" . . 91

Figure 2-11 Spatial distribution of households by
parish, for the asthma study . .. 93


Figure 3-1 Inter-migration patterns by parish,
1970-1990 . . 133

Figure 3-2 Settlement and Land Use Policy of
Barbados . .... 135

Figure 3-3 Chattel house floorplan ... 165

Figure 3-4 Sketches of the three stages of the
chattel house . .. 166

Figure 4-1 Correlation matrix (p-values) for structural
and demographic variables in the Asthma Study
households . . 239

Figure 4-2 Correlation matrix for structural and
demographic variables in the asthma study
households . . 240

Figure 4-3. The Allergy Disease Model in the
Architectural Perspective . .. 241

Figure 6-1 Graded Skin Test Response (Krouse &
Klaustermeyer 1980) . .. 344

Figure 6-2 Correlation matrix for pest allergies by
structural and demographic variables 354

Figure 6-3 P-values for pest allergies by
structural and demographic variables .. 355

Figure 6-4 Correlation matrix for positive skin
tests, indicating the interrelationship between
each of the extracts . .. 356

Figure 7-1 Modelling the variables: a summary 384


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Kathleen Carole Barnes

December, 1992

Chairperson: Dr. Gerald Murray
Major Department: Anthropology

Allergy is referred to as part of the epidemiological

transition of disease; it is uncommon in traditional

societies, is steadily increasing in developed societies, and

demonstrates dramatic incidences in transitional societies.

Some of the most antigenic allergens are from arthropods

(arboallergens). Pan allergy to insects and noninsect

arthropods has been exhibited in individuals who have

previously been sensitized to only a few arboallergens.

Previous research demonstrated that D_. teronvssinus (pp)

was the predominant house dust mite (HDM) in Barbados and that

wood homes on rock foundations provided an ideal mite habitat.

The contemporary trend is an integration of permanent features

into the traditional house form (e.g., concrete and indoor

plumbing into the traditional house form); this trend is a


metaphor for sociocultural change because of the historically

tenuous nature of land tenure and housing in Barbados.

Concrete, plumbing, and other "modern" features have been

incriminated in increasing indoor relative humidity and

reducing air flow, two factors which create favorable habitats

for a number of household pests.

Asthma is increasing in Barbados; asthmatic attacks

reported to the Casualty Clinic doubled in 10 years to 7,137

in 1990. The current study was undertaken to examine the

effects of contemporary housing on the distribution of

household pests and the harmful effects that exposure to

arboallergens might have on the asthmatic. Densities of pg

were found in 100% of homes sampled and were higher in

concrete homes (f=9.55, p<0.001); qualitative data suggest

that other pests prevail in both types of construction.

Asthmatic children (168) were skin tested against 11 household

pest allergens; HDM elicited the greatest number of responses

(77.4% to DD and 74.4% to D. farinae, or Df). Allergy to P2

was correlated to living in a concrete home (X2=4.37, p=0.04).

The high incidence of allergy to Df despite low or absent

densities and a positive reaction to an average of four

different pests suggest cross-reactivity. The implications of

the study are the following: housing trends have a direct

influence on the presence of Dp, a highly allergenic pest; and

sensitivity to DD possibly predisposes asthmatics to develop

other arboallergen hypersensitivities, thus increasing the

likelihood of morbidity related to arboallergens.



Introduction: Rationale for the Study

The Relationship Between Modernization and Allercg

As humans have evolved culturally, we have increasingly

exposed ourselves to numerous pathogens due to our

alterations in the environment, complex food-producing

practices, and intense contact with other infected humans

and their waste products (for many populations this picture

remains the same today, although the sociopolitical reasons

have changed with time). The bulk of these pathogens in

earlier agriculturalist populations were those responsible

for infectious disease, usually vector-borne and parasitic

and not dependent on large populations or even human hosts

for their survival. With the development of industrial

societies, cultural advances in technology enhanced the

provision of better sanitary practices (e.g., systems for

waste disposal, treated water), which eliminated a number of

the parasitic diseases, and biomedical advances provided

chemotherapy and immunization against many of the

debilitating and mortal communicable diseases. As a result,

large numbers of people living in what are considered

"developed" countries have never even been exposed to a


parasite or to many of the viruses and bacteria that

eliminated whole societies. This process has resulted in

what has been termed the "epidemiological transition" of

disease, defined as a trend from acute infectious diseases

to chronic noninfectious disease and generally industrial

and degenerative in nature (Omran 1971; Corruccini and Kaul

1983). The increasing prevalence of these diseases is

related to human modernization and urbanization.

Extreme changes in contemporary modes of production

have resulted in the introduction of new organic and

inorganic substances in the contemporary macroenvironment.

These substances are inhaled and ingested both at home and

at the work place, or the microenvironment. There is

extreme class consciousness, as well as continuous efforts

to "move up" socially and "improve" the home environment

with the addition of non-consumable goods and climate

control. The net result of these cultural changes is

exposure to large amounts of substances never before

encountered in human history and, more importantly,

subjection to these substances in a very short period of


One of the new chronic diseases resulting from the

epidemiological transition is allergy. Although we do not

know the degree of prevalence of allergy in previous

populations, we do know that the incidence of allergy is on

the rise (Corruccini & Kaul 1983; Massicot & Cohen 1986;

National Asthma Education Program 1991). Contributing to

this is exposure to many antigens that, for reasons not

completely understood, are highly allergenic. The most

common of these are pollens, molds, and notably, arthropods

(Gergen et al. 1987). Schwartz (1990) argues that

arthropods are more important as allergens than are pollens

because they are so widespread.

The Role of Household Pests

Possibly the most significant impact in regard to

allergic disease as a result of modernizing our

microenvironment has been the increased exposure to peri-

domestic and domestic arthropods and rodents. In

elaborating the home environment for the purpose of removing

ourselves from climatic elements, we have provided ideal

habitats for a number of "pests." This is most evident in

the study of urban entomology; household infestations of

most pests increase with the degree of urbanization (Aedes

SPP. mosquitoes, Gratz 1973, Slosek 1986; cockroaches,

Ebeling 1975, Koehler et al. 1987; rodents, McNeill 1976).

What is most alarming about the relationship between

urbanization and household pest infestations is the extreme

allergenicity of arthropods and certain other animals.

The relationship between an increase in the incidence

of allergy in general and modernization has been largely

attributed to the development of housing conditions that are

favorable to house dust mites (Massicot & Cohen 1986). In

fact, the most predominant allergen among residents in

developed and developing countries alike is the house dust

mite (Buchanan & Jones 1972; Kang & Sulit 1978; Dowse et al.

1985; Massicot & Cohen 1986; National Asthma Education

Program 1991). Dowse et al. (1985) studied the incidence of

asthma in eight South Fore villages in the Eastern Highlands

of Papua New Guinea. There had been an alarming increase in

prevalence over a 10-year period; the overall rate was 39

and 46 times higher than rates noted in 1972 and 1975,

respectively, and was much higher than rates reported from

other highland regions. The reason was attributed to the

introduction of woolen blankets into the villages;

Dermatophagoides pteronyssinus and D. farinae were found in

97% of the blankets used by the Fore (Dowse et al. 1985).

Baldo and Panzani (1988) suggest that "pan allergy" to

insects may exist in individuals who have previously been

sensitized to one or several insects and that allergenic

similarities possibly extend to other noninsect arthropods,

such as spiders or crustaceans. The ramifications of this

finding are frightening from an epidemiological point of

view when one considers the widespread distribution of

arthropods, and the consumption of many of those arthropods

(e.g., shrimp, lobster). It implies that any household pest

is a potential source of allergen.

Numerous studies have been published that confirm the

relationship between cockroach allergens and bronchial

asthma (Kang 1976; Kang & Sulit 1978; Kang & Chang 1985;

Kang 1990). Individuals of lower socioeconomic status

within the United States reportedly have higher incidences

of cockroach hypersensitivity compared to individuals of

higher socioeconomic status (Bernton & Brown 1967; Kang &

Chang 1985). Authors conclude that this likely is

correlated to higher cockroach infestations in poor housing

and related to the duration of exposure (Bernton & Brown

1967; Twarog et al. 1977). However, the reverse is true in

developing countries (Marchand 1966; Brenner et al. 1990;

Brenner et al. 1991). In Puerto Rico, Marchand found that

his patients who demonstrated hypersensitivity to

cockroaches were from the middle to upper socioeconomic

strata (1966). In the Dominican Republic, the incidence of

allergy to cockroaches was proportionately related to the

quality of the house structure (X2=7.36, df=2, p<<0.01);

individuals from higher quality homes had a greater

incidence of allergy to cockroaches (Brenner et al. 1990;

Brenner et al. 1991). Data suggested that the higher

quality homes built of masonry had less air flow than the

homes of wood, thereby producing a more favorable

environment for the cockroach and also an accumulation of

arboallergens.' More favorable environments and larger

populations also may increase the likelihood of physical

contact with cockroaches, resulting in an injectant or

ingestant contact with the allergen; cockroach "biting" was

reported in heavily infested, urban environments in both

North Central Florida and the Dominican Republic (Barnes &

Brenner, in prep.).

In another study, a relationship was found between the

presence of house dust mites and the structure of houses.

Pearson and Cunnington (1973) identified 23 species of house

dust mites in Barbados (D. Dteronvssinus was the most

predominant) and concluded that they were so proliferative

because of the timber dwellings situated on top of porous

coral rock foundations, which, close to the soil, generated

a sufficient amount of moisture to provide excellent

habitats for the house dust mites.

Objectives of the Study

Based on the concept that allergy is a disease of

modernization and arboallergens as well as other pest

allergens (e.g., rodents) play a significant role in the

incidence of allergy, a study was designed to test the

hypothesis that modernization of the domestic environment is

related to allergy to household pests. The domestic

environment was chosen as the focal domain because two of

the best-known producers of arboallergens--the house dust

mite and the cockroach--are domestic pests. Certainly the

most concentrated human exposure to arthropods is indoors

rather than outdoors.

As with most other diseases that are part of the

epidemiological transition, there is a strong genetic

component in the development of allergy. Although people do

not have control over their genetic make-up, they do have

control over their behavior and to a limited degree, control

over their environment. Behaviors and the environment over

which the lay person probably has the most control are those

related to the home. Because "household pests," as the

definition connotes, thrive within the domestic domain,

certain sociodemographic variables related to the home

likely play a role in both household pest infestations and

allergy to pests. If such variables can be controlled by

the people living in the affected environment, then

residents might be able to play a central role in their -own

promotion of health and well-being.

Herein lies the guiding research question: What

existing human perceptions, behaviors, and household

environmental factors contribute to the notion that allergy

is a disease of modernization? Specifically, are household

pest infestations influenced by modernization of the home,

and to what degree do they contribute to allergy? From an

immunological perspective, it is suspected that the

relationship might be due to the likelihood and degree of

exposure to certain household pest species. Exposure to

household pest allergens primarily is dependent on two

factors: first, there must be a substantial population of

the pest for it to produce allergens (most typically

aeroallergens); and second, the domestic environment must be

conducive to the maintenance of the pest population and/or

the accumulation of the attendant allergens. It is apparent

that modernization of specific features of the home

increases the degree of pest infestations indirectly, by

providing a suitable habitat for the pest. Additionally,

cultural and socioeconomic characteristics unique to a given

population may accelerate the development of certain

architectural and sociodemographic features. The primary

objective of the following study was to identify these

features and to create a model for predicting the likelihood

of developing allergies to certain pests in the event that

these features are present.

The following paragraphs present an elaborated view of

the epidemiological transition of disease. Following this

presentation, the evolutionary perspective of allergy will

be discussed, so as to better understand the

interrelationship between the physiological and cultural

elements of allergy.

Defining the Epidemiological Transition

An interesting characteristic of many of the

contemporary, chronic diseases is that they are particularly

prevalent and "epidemic"-like in transitional societies, or

those populations undergoing the shift from developing to

developed modes of production. In developing countries,

many of the chronic diseases associated with the

epidemiological transition appear first in members of the

upper socioeconomic strata (Burkitt 1973), probably because

of their access to Western products and practices. The

diseases typically share common, etiological factors related

to human culture, including diet, activity level, mental

stress, behavioral practices, and environmental pollution.

There are a number of contemporary chronic diseases

that may be used to exemplify the epidemiological transition

model. Biocultural anthropologists have investigated

dietary changes as an important etiological factor for the

development of many chronic diseases. Burkitt (1973) notes

that while the intake of fiber in British and American diets

between 1880 and 1960 fell by more than 90%, fat consumption

increased, and sugar consumption doubled. A low fiber, high

fat diet seems to have the greatest effect on the

development of noninfective bowel disease, including

diverticular disease, appendicitis, colon cancer, polyps,

and ulcerative colitis (Burkitt 1973). Similarly, high

serum cholesterol has been widely discussed as a major

factor in coronary heart disease (the number one cause of

death in developed countries) and gall-bladder disease

(Burkitt 1973). Obesity and high intakes of refined

carbohydrates are related to the increasing incidence of

heart disease and diabetes. Diabetes is particularly high

in groups undergoing the transition from traditional life-

style to "modern"; certain urban groups of the Bantu in

South Africa have rates 40 times greater than their

counterparts in the rural sector (Pelto & Pelto 1983).

Similarly, many Amerindian groups, including Mexicans,

experience unusually high incidences of obesity, adult onset

diabetes mellitus, and gallbladder disease; this has been

termed the "New World Syndrome" of disease (Weiss et al.

1984). Obesity is considered to be the most common form of

malnutrition in developed countries (Burkitt 1973; Pelto &

Pelto 1983) and is a direct result of an increasingly

sedentary life-style (which reduces caloric needs) in

conjunction with steady or increasing caloric intakes.

Medical technology has succeeded in eliminating many of

the communicable diseases that reduced whole populations

(e.g., the Black Death in Europe, smallpox in the Americas),

and chemotherapies and immunizations have the potential to

control a majority of other infectious diseases. It has

been calculated that public health measures alone (e.g.,

clean water and sewage provisions) increase life expectancy

in a population to almost 65 years (Baker 1989). The

control of infectious and communicable disease, resulting in

increasing life expectancies, has paved the way for

degenerative disease.

A feedback effect of medical technology is a shift in

demographic trends, whereby the rate of fertility continues

at a steady state while mortality drops sharply. The

eventual matching of fertility rates with mortality rates is

referred to as the "demographic transition," and this is the

current pattern in most of the developed countries (Baker

1989). However, societies in transition are slower to

achieve the equalizing of fertility with mortality, probably

largely due to the suddenness of changes in mortality rates.

The continuing increase in population size in conjunction

with an increase in commercialization of agriculture--

characterized by a reduced need for workers in the rural

sector and malnutrition--has had particularly detrimental

effects on the urban environment in transitional societies.

These urban environments are characterized by over-crowded

living arrangements, underemployment or unemployment, and a

chronic state of infectious disease (e.g., tuberculosis,

filariasis) for which biomedical treatments are not always

available (Miller 1973), plus the introduction of many of

the chronic, "modern" diseases.

While degenerative diseases are characteristically

higher in persons of older age, they are not limited to that

population. Baker (1989) argues that degenerative diseases

are not necessarily an inevitable process of ageing, because

many of them (e.g., myopia, high-frequency hearing loss in

middle-age) are completely absent in some groups.

Similarly, while malignancies can be found in all human

populations, and for that matter, other animals, the types

and locations of malignancies vary from traditional

populations to "modern" populations (Baker 1989).

A unique characteristic of the chronic diseases is

their relatively recent appearance in human history as a

major cause of morbidity. According to Corruccini and Kaul

(1983), this is indicative of a strong environmental factor

in disease etiology. Indeed, all of the above disease-

causing factors are those created by humans and their

culture. While biological factors such as genetics are no

doubt important in determining who is most likely to succumb

to which disease, genetics alone cannot explain the rapid

spread of chronic disease. Our genetic systems simply do

not operate in selecting for or against certain

physiological features in such a short period of time.

When evaluating the epidemiology of a specific disease,

what we see in developed countries is the end result of the

epidemiological transition; at this level it is difficult to

identify specific socio-environmental variables that are

responsible for the disease process. By including in the

epidemiological analysis less-developed societies and their

socio-environmental variables, we create the "whole

picture," or the evolutionary perspective of that disease.

By analyzing the natural history of a disease in different

societies and the different human behaviors associated with

the distribution of that disease, we will be able to

identify disease-causing variables and predict

epidemiological outcomes. This information can then be

applied to health education, and preventive medicine may, in

the ideal setting, be used to minimize the effects of that


Allergy is an excellent example of a disease which is

part of the epidemiological transition. As heretofore

described, allergy is uncommon in traditional societies,

reaches epidemic rates in transitional groups, and is

increasing in developed populations where it is most common

overall. Prevalence of allergy is characterized by the

degree of modernization within a population, for reasons

similar to--and in some ways quite different from--the

typical chronic diseases.

Disease Ecoloqg and the Evolution of Human Allergy

The theoretical framework most commonly used in

understanding the evolution of human disease is based on the

triad of host, pathogen, and environment. The current

ecological perspective places the most emphasis on the

environment and gives equal attention to both the physical

components (e.g., geologic, climatic, biotic) and the

cultural components (e.g., technological, social, and

ideological). The host represents both the individual and

the population and is responsive and constantly rallying to

maintain homeostasis. The pathogen is any insult against

the host.

The evolution of human disease has closely paralleled

the evolution of human modes of subsistence practice, and

this is probably why so many biocultural anthropologists


have referred to the different subsistence periods in human

history as a construct for analyzing disease evolution

(Cockburn 1971; Dunn 1977; Armelagos & Dewey 1978; Baker

1989; Armelagos et al. 1990). Subsistence modes dictate, to

a degree, biocultural factors such as demographics and

dietary practices.

Disease in Pre-State Society

Humans and their predecessors relied on hunting-

gathering as a means of subsistence from about four million

years ago until the beginning of the neolithic, some 11,000

years ago. Hunter-gatherers utilize as many edible products

as possible, including a variety of protein and carbohydrate

sources as food, and have well-balanced diets. (There are

exceptions; for example, Arctic aboriginals have evolved

physiological and cultural mechanisms whereby they survive

on a diet with little variation, mostly composed of animal

protein.) Hunter-gatherers are mobile for the purpose of

finding new sources of food after a particular region has

been exhausted and are small in population size to

facilitate that mobility. Their mobility negates the

possession of many nonconsumable goods; therefore, their

living quarters are simple and constructed of the ecological

materials readily available to them. Because of the large

area of land needed to support their subsistence, there is

limited contact with populations outside of the immediate


Due to the small population size of early hunter-

gatherers, there were a limited number of infectious

diseases that could have established a symbiotic

relationship with human hosts. This factor highlights

several epidemiological principles essential for

understanding the evolution of human disease. Pathogens (a

collective term used here to represent bacteria, viruses,

fungi, spirochetes, and the parasites, including protozoa,

helminths and arthropods) must, in order to survive, affect

low mortality among their hosts. An underlying principle

when determining the evolutionary association between a

pathogen and a host is that the more stable the host-

pathogen relationship, the longer the two likely have

coexisted. In contrast, more pathogenic relationships are

assumed to have been recently developed, because they

apparently have not evolved a stable relationship. The

elaborate relationship that helminths have developed with

human hosts is indicative of their long history with

hominoids. For example, no single helminth parasite is

limited to humans as a necessary host, and all of them have

avoided the development of complete immunity in the human

host. This lack of immunity probably is indicative of the

very low mortality caused by helminths and complete immunity

is therefore not essential for the host's survival (Kliks

1983). An exception to this principle are pathogens that

cause high mortality--such as tetanus--but do not require

the host for their survival.

Another epidemiological principle is that the more

pathogenic the insult by the organism, especially against

hosts of a pre-reproductive age, the greater the selective

pressure will be on the population (Kliks 1983). Diseases

that produce low pathogenicity or are chronic and therefore

affect the post-reproductive population will exert less

selective pressure. Physiological adaptation (e.g., in the

form of acclimation, accommodation, or acclimatization) is a

response to selective pressures by which a host improves

his/her chance to rally against pathogenic insults, and may

take the form of genetic changes or physiological responses

completed within a single life span. Cultural "adaptation"

includes responses by which information is transferred to

subsequent generations to enhance survival in a given

environment. Genetic changes are the slowest adaptations to

evolve and signify long-term coexistence.

Many of the infectious diseases require large host

populations and rapid transmission time for their survival.

This is referred to as the population "threshold level"

(Cockburn 1971). If a host population is not a minimal size

for pathogen maintenance, then the pathogen will die.

Typically with infectious diseases large host populations

are required; the acute communicable infections of cholera,

rubella, smallpox, mumps, and measles are examples. The

population threshold is not necessarily due to mortality,

because even pathogens that can live as commensals within

the host often will die in a small community (Cockburn

1971). Furthermore, organisms that are highly pathogenic

often elicit an immune response, that subsequently limits

the number of susceptibles in a small population.

Two types of infectious diseases have been recognized

as affecting early hunter-gatherers: those which had

existed among the prehominids and persisted in the hominids

and those that were accidentally encountered from other

animals zoonosess) by means of vector, wounds, or

consumption of animals (Polgar 1964). Cockburn (1971)

listed a number of intestinal parasites, ectoparasites,

treponemal infections (including yaws), malaria, and several

viruses (herpes and hepatitis) that were present in primates

prior to the first humans. Many of the zoonoses were

acquired after eating infected or raw mammals, insects,

birds, and fish, the most common of which were probably

anthrax and botulism (Cockburn 1971). Zoonotic diseases

also were spread by arthropod vectors. Avian, or ichthyic,

tuberculosis was a disease of early humans (Cockburn 1971),

as was schistosomiasis (Baker 1989).

Hunter-gatherers rarely experienced malnutrition or

starvation, and chronic disease was infrequent (Dunn 1977).

Nevertheless, life expectancy was relatively short, rarely

longer than 30-35 years, with few living longer than 50

years (Baker 1989). It is believed that traumatic and

accidental death was the leading cause of mortality,

including falls, snakebites, and death by predators. Social

mortality involving infanticide, homicide, suicide, war, and

cannibalism was also a factor in mortality (Dunn 1977; Baker


Disease in State Society

The shift to agricultural forms of subsistence some

10,000 years ago marked some of the most significant changes

in culture since humankind's emergence. Large-scale

production of food changed the nomadic life-style of the

hunter-gatherer to a sedentary one. This resulted in the

accumulation of goods and the advancement of technology,

thus promoting the division of class and trading and

communication with other populations. To plant and harvest

farms and to replace animal sources of protein lost after

clearing land for the farms, many groups domesticated and

herded animals.

All of the above either directly or indirectly

contributed to human disease, most notably infectious

disease. The concentration of large numbers of human hosts

in a permanent location provided the "threshold" population

level necessary for the communicable diseases as well as the

vector-borne diseases. The sedentary life-style resulted in

a build-up of human waste proximal to living quarters and

water supplies, which promoted disease from parasites such

as the ascarids and hookworms. Contact with other

populations facilitated the spread of infectious disease to

epidemic proportions by providing organisms with "virgin"

hosts. New zoonotic diseases were encountered with the

constant contact of domestic animals, including anthrax,

brucellosis, tuberculosis, and Q fever (Polgar 1964).

Unwanted peri-domestic animals such as rodents and sparrows

developed permanent habitats in and around human dwellings.

By 3000 B.C. cities with populations of 50,000 or

greater were established in the Near East (Armelagos et al.

1990). All of the characteristics of the earlier

agricultural populations were present, but on a much larger

scale. There were increasing difficulties with supplying

water and food and with managing human waste, resulting in

outbreaks of cholera (Armelagos & Dewey 1978). Increasingly

complex societies also resulted in a breakdown of

traditional social practices; one example is the increase in

sexually transmitted diseases as a result of sexual

promiscuity (Armelagos & Dewey 1978). The same communicable

diseases that affected earlier agriculturalists continued,

but for the first time the populations were sufficient for

maintaining an endemic form (Armelagos et al. 1990). Yet

what was endemic to one population was often detrimental to

another. Cross-continental trade and exploration resulted

in intense epidemics (Zinsser 1936; McNeill 1976; Laird

1989). Large-scale wars resulted in some of the most

serious epidemics in history. Following the introduction of

Rattus into Western Europe by the Crusaders and the

introduction of the Xenopsvlla cheonis flea by seaborne

vessels, the Black Death began to take its toll in Europe in

1347 and, three years later, had eliminated at least a

quarter of the European population (approximately 25 million

people) (Laird 1989).

With increasing developments in technology evolved the

germ theory. A better understanding of disease causation

has admittedly resulted in increasing control over

infectious diseases. With the discovery of immunization

came the eradication of smallpox, and most of the other

communicable diseases have diminished in distribution. The

decrease in infectious disease has resulted in greater life

expectancies, so that chronic, degenerative diseases have

become the focus of morbidity and mortality. It has also

resulted in higher fertility rates producing worldwide

population explosions and leading to overcrowding,

underemployment and unemployment, and subsequently all of

the diseases associated with poverty.

Recently, much attention has been focused on the

detrimental effects of industrialization on the

international environment, including water, land, and

atmosphere. Massive industrial production of commodities

has caused pollution of much more than human waste.

Increasingly, there is concern over the health implications

of contaminated water supplies, overuse of pesticides in

commercialized agriculture, atmospheric chemicals, and the

future effects of a depleted ozone on human health and food

production. Increasing incidences of cancer among young

people and the increase in respiratory disease has been

implicated in these environmental changes. While most of

our current chronic diseases and conditions most certainly

existed in early populations (e.g., dental malocclusion,

myopia, and allergy) they must have been rare in order for

those populations to have survived without the technological

compensations which we know today, and their relatively

sudden appearance on such a large scale suggests an

environmental causation (Burkitt 1973; Corruccini & Kaul


Application of the Disease Ecology Model: Allergy

Clearly, human biological and cultural activities

interacting with the environment forged lasting yet dynamic

relationships between host and pathogen. However, the

effects are two-way. Until this point I have focused on how

human evolution has influenced disease, but at least as

important is the issue of how disease has influenced human

evolution. This has stimulated countless biological and

cultural adaptations so that the host--individual or

population--may be returned to a steady state of well-being.

In many instances these adaptations have feedback effects on

the disease ecology.

The effect of disease on human evolution can be

summarized as follows: as humans evolved from one stage of

subsistence to another, they increasingly altered their

environment, thereby promoting new diseases which did not

exert selective pressures on previous populations to the

same degree. These newly encountered diseases forced

biological and cultural adaptations on the population,

compelling society to alter or increase food production,

explore new territories for new resources, and advance

technologically to combat the insults. Because no two

environments or no two gene pools are identical, different

diseases evolved for different populations. With increasing

communication between groups, there was an increase in the

distribution of disease. This "evolution" is circular;

disease forced changes in the host and her/his environment,

changes were made, and new diseases arose, forcing more


The evolution of human allergy is an excellent

illustration of this feedback effect. Allergy is an example

of how a disease response (hypersensitivity) possibly

evolved as an adaptation to another disease early in human

history and how this "adaptation" is currently causing

negative feedback effects on contemporary disease ecology.

The Influence of Disease on Biological Adaptation:
Parasitic Infestations and IqE

At the beginning of this essay it was noted that the

earliest diseases affecting humans were those which had

affected the prehominids and zoonoses. It was also noted

that one of the "epidemiological principles" is that the

more stable the host-pathogen relationship, the longer the

two have probably coexisted. Helminths were illustrated to

have evolved this type of relationship with human hosts, and

the very long history of that relationship and low mortality

are evident in that all of the helminths have avoided the

development of complete immunity in the human host,

indicating that complete immunity is not an essential

feature for the host's survival (Kliks 1983).

Given that parasites and humans coexisted for such a

long period, one would assume that the disease might have

influenced some sort of biologically adaptive means for

minimizing the pathogenic effects in the human host. We

already know, for example, some of these adaptive

mechanisms. Many of the parasites have evolved mechanisms

for assuring the survival of the host, as in concomitant

immunity, in which the adult worms prevent the survival of

subsequently acquired larvae of similar or different species

(Kliks 1983). The effects of more elaborate genetic

adaptations have also evolved (e.g., sickle cell trait),

which are specific to a parasitic species. Researchers

examining contemporary populations chronically infested with

parasites observe another common feature: stimuli that

cause some of the highest titers of serum IgE are the

protozoa and helminths (Johansson et al. 1968; Ito et al.

1972; Bazaral et al. 1973; Desowitz 1981; Sher & Ottesen

1988). Yet, while serum IgE is 10 times higher in people

with allergy than in normal people, IgE is 10 times higher

in people with parasitic infections than it is in allergic

(Sher & Ottesen 1988).

There are five immunoglobulins in the human immune

system: they are, in order of concentration, IgG, IgA, IgM,

IgD, and IgE. Immunoglobulins are more generally referred

to as antibodies and are produced from lymphocytes. Of all

the human immunoglobulins, the least is known about

immunoglobulin "E," referred to as IgE. Elevated IgE is

found only in two clinical conditions: helminthic

infestations and allergy. Allergy is the common name for

"hypersensitivity," or an inappropriate response to a

harmless antigen upon second contact with that antigen.

There are four types of allergic reactions, but "allergy"

generally refers to the Type I reaction, also called

immediate type hypersensitivity. The main mediator in Type

I hypersensitivity is IgE.

IgE is dispersed both free in the serum and bound on

the surface membrane of mast cells and basophils. Serum IgE

is very low in comparison to concentrations of other

immunoglobulins, typically ranging from 10 to 100

nanograms/ml, or, put another way, is less than 0.00001 of

all the body's total immunoglobulins (Johansson & Bennich

1985). IgE is unique among the classes of immunoglobulins;

although its normal concentration is extremely low, it may

increase several hundredfold following challenge to a

specific stimuli (Barbee et al. 1981).

Allergenic stimuli are referred to as allergens or

antigens. Antigens may be inhaled, injected, ingested, or

presented by dermal contact. When IgE comes into contact

with an antigen, it triggers the release of the contents in

the basophils and mast cells. These mediator substances

include: histamine, a vasodilator; slow-reacting substance

of anaphylaxis (SRS-A), a smooth muscle contractor; and the

eosinophil chemotactic factor of anaphylaxis (ECF-A), which

causes the accumulation of eosinophils where the interaction

occurs. The clinical feature of Type I hypersensitivity is

referred to as "atopy." Atopy can be manifested as asthma,

allergic rhinitis, or eczema, sometimes called dermatitis

(not to be confused with "contact dermatitis" which is a

manifestation of Type IV hypersensitivity). Atopic

individuals are, in the most extreme scenario, at risk of

anaphylaxis, a life-threatening reaction in which there is

vasodilation and constriction of smooth muscles,

particularly of the bronchus, thus interrupting the exchange

of metabolic gases.

The capacity for developing elevated levels of serum

IgE is largely determined by genetics. This predisposition

is controlled by the major histocompatibility complex, or

MHC (antigen-specific) and by a regulator gene (nonspecific)

(Menser et al. 1975; Marsh et al. 1981; Marsh et al. 1980b).

Prevalence of atopy is higher in non-European descendants

living in developed countries (Davis et al. 1961; Worth

1962; Orgel et al. 1974; Marsh et al. 1980a; Waite 1980),

While the functions of IgG, IgA, IgM and IgD are

relatively well understood, the physiological function of

IgE is less clear. IgE is important in facilitating fluid

transport across cell membranes, and it also functions in

fighting bacterial and viral disease, although to a much

lesser degree than the other immunoglobulins (Gerrard 1985;

Johansson & Bennich 1985). With increasing evidence of a

relationship between elevated IgE and parasitic

infestations, researchers began to determine what the

physiological relationship was between the parasite and the


Animal studies demonstrated that serum IgE appeared to

function by minimizing the number of parasites in a host who

was chronically exposed to large numbers of parasites (Hsu

et al. 1974; Dessein et al. 1981). Dessein et al. (1981)

concluded that IgE-suppressed rats demonstrated

significantly less resistance to infection by Trichinella

spiralis than controls with normal IgE levels. Following

sequential challenges with Schistosomula iaponicum cercariae

to a rhesus monkey, hypersensitivity reactions were

determined using microscopic examination of skin biopsies

and macroscopic determinations wheall and flare) (Hsu et al.

1974). The schistosomulae were destroyed in the dermis of

the primate approximately 12 hours after the challenge.

After numerous other laboratory and clinical studies,

contemporary immunologists concluded the following

protective mechanism of IgE against heavy parasitic

infections: upon entering the host, soluble parasitic

antigens diffuse across the intestinal mucosa and are

transported to the lymph nodes, where an IgE-mediated

response occurs. Mast cells migrate to the same lymph

nodes, are sensitized by the development of parasite-

specific IgE on their surface, and return to the intestinal

mucosa. Upon contact with the parasitic antigen, the mast

cells degranulate and release their mediators, which

subsequently attract to the site eosinophils, complement,

and parasite-specific IgG, all of which function to damage

and expulse the parasite (see Brothwell 1972).

Having reviewed the relationship between parasitic

infections and IgE, and with the understanding that the only

other manifestation involving elevated IgE is allergy, the

question is posed: is there any relationship between

parasitic infections and allergy? Unfortunately, this

relationship is much less clear than the helminthic-IgE

relationship. Before discussing what is known about this

topic, a review of the evolutionary history of IgE as an

adaptive process will be presented.

The Influence of Disease on Cultural Adaptation: The
Gradual Decrease in Exposure to Parasites

With the confirmation of a relationship between IgE and

both allergy and parasitic infestations, there have been a

number of suggestions as to a "cause-and-effect"

relationship between the two. From an epidemiological point

of view, this information would be valuable in determining

the distribution of IgE-related disease in populations of

the past and predicting trends in contemporary societies.

The first evidence of a relationship between IgE and

parasites ironically came from a temperate environment

rather than the tropics, where parasitic infections are

typically endemic. D. Tullis (1970) declared that an unusual

epidemic of Ascaris lumbricoides, Trichuris trichiura,

Necator americanus, and Ancylostoma duodenale in the Niagara

Falls vicinity had coincided with an alarming increase in

the prevalence of asthma. One-hundred and ninety-eight of

the 201 asthmatics were diagnosed with intestinal

infestations involving one or several of the above

parasites. Tullis' conclusion was that there was a definite

correlation between the development of bronchial asthma and

intestinal parasitic infestations. Following Tullis'

report, other immunologists published similar findings,

concluding that parasites caused allergy (Huntley 1976;

Joubert et al. 1980; Desowitz et al. 1981).

The investigation was taken to the tropics where

populations traditionally have experienced high levels of

parasitic infections. As expected, and similarly to the

previous studies, populations that were heavily infested

with parasites demonstrated extremely high levels of IgE

(Godfrey 1975; Warrell et al. 1975). However, one

researcher, Godfrey (1975), noted in his study that, among

those patients with the extremely high serum IgE and

parasitosis, allergy was practically nonexistent.

Furthermore, he added a socio-demographic variable,

concluding that parasitosis was highest in the rural region

and asthma was highest in the urban region, where there was

a low incidence of parasitosis. Similar experiments in

laboratory settings revealed the same results; allergy was

minimized or absent in persons with heavy parasite

infections (Phills et al. 1972; Hsu et al. 1974; Godfrey &

Gradidge 1976; Dessein et al. 1981).

Curiously, other studies indicated that there was no

relationship between IgE, parasitic infections, and allergy;

atopy was just as prevalent in persons with helminthic

infestations as without them (Alcasid et al. 1973;

MacFarlane et al. 1979). These studies did not, however,

take into consideration various socioeconomic and

demographic variables, such as urban or rural residence.

The prevailing conclusion is that, under certain

conditions, all three of the above scenarios may occur. But

in the majority of cases involving heavy parasitosis in

endemic regions, the incidence of allergy to both parasites

and exogenous allergens is low. The mechanism for that

suppression is believed to be one of the following two

theories. Chronic exposure to large numbers of parasites

stimulates the production of both specific and nonspecific

IgE antibodies, resulting in "saturation" of the IgE

receptors on the mast cells and therefore preventing the

cell from responding to additional "specific" antigens, such

as molds or arthropods (Godfrey & Gradidge 1976; Ottesen


A second--and currently more widely accepted theory of

mechanism--is a "blocking antibody" mechanism, in which

persons chronically exposed to helminths produce antigen-

specific IgG-blocking antibodies which inhibit to a degree

the mast cell or basophil degranulation process (Ottesen

1985). Similar to other mechanisms involving human hosts

and parasites, the potential role of these antibodies

appears to be more to limit the degree of hypersensitivity

rather than to eliminate the reaction completely (Ottesen et

al. 1981; Ottesen 1985). This mechanism is specific for

maintaining a low incidence of allergy to parasites rather

than exogenous allergens and is not recognized in atopic

disease (Ottesen et al. 1981). It is, interestingly, also


the principle for which desensitization therapy for allergic

disease is administered (Ottesen et al. 1981; Ottesen 1985).

The blocking antibody theory explains why allergy to

helminths usually is seen only in the early phases of

parasitic infestation. At the sites of initial parasitic

contact, such as the mucous membrane or skin, the antigen

concentration is too great to succumb to the effects of the

blocking antibodies, yet as the parasitosis progresses and

sufficient IgG-blocking antibodies are generated, the

hypersensitive reaction gradually would be controlled

(Ottesen 1985). Consequently, in regions where

helminthiasis is in an acute, epidemic stage (e.g., Niagara

Falls) allergic responses to the helminths would be


A blocking mechanism specific for helminths--but not

for other allergens--would also explain why helminthiasis

and allergy occasionally occur simultaneously. Populations

that have been chronically exposed to parasites have a

genetic predisposition for elevated IgE. It is likely that

the occurrence of both phenomena is reflective of

populations undergoing a transition from developing to

developed living conditions, as in rural to urban migrations

in tropical countries, where they are exposed to both

parasites and new and numerous exogenous allergens.

Clearly, there is a cause-and-effect mechanism between

IgE and both helminthiasis and atopy. However, the sum

process is multi-causal. First, predisposition to develop

high levels of serum IgE is genetically-determined. The

fact that these elevated levels are higher in non-Europeans

than Europeans has already been mentioned. Presumably, the

selective advantage for a haplotype which is predisposed to

producing high levels of IgE--a key component the mechanism

for destroying invading helminths--would be highest in a

population chronically exposed to helminths. This

represents the protective role of IgE. Over millennia, and

following the cultural impact of widespread distribution of

once isolated gene pools, these haplotypes would also be

widely dispersed. Elevated IgE in populations chronically

exposed to parasites apparently did not produce adverse

health effects due to the check-and-balance mechanism of

IgG-blocking antibody and the control of the hypersensitive

state. This mechanism was advantageous to both the host and

pathogen. Without exposure to helminths, however, there is

apparently no IgG-blocking antibody and therefore no built-

in control over the immune response in the event of contact

with foreign allergens. This is the negative feedback

component of the immune mechanism. Traditional societies

exposed to many parasites--but few exogenous allergens in a

less complex environment--benefitted from the biological

adaptation of IgE; parasitosis was kept at a reasonable

state of low morbidity. Yet as humans evolved culturally

and biologically, they created an environment in which IgE

is not checked (e.g., no exposure to parasites) and in which

potential allergens abound.

In summary, it is apparent from this particular

illustration--IgE, helminths, and parasites--that disease

plays a major role in the adaptation of the human host, but

that many adaptations may in turn promote additional

diseases. A new trend in the ecological perspective of

disease is to shift our focus on adaptation as an "optimal"

mechanism to one of sufficiency. Gould and Vrba (1982)

argue that, in the case of evolution, the word "adaptation"

has been overused, because not all features that enhance

fitness were necessarily designed by natural selection for

their contemporary role. The authors suggest that

"exaptation" be used to explain certain characters that

either evolved for other purposes or for no purpose at all,

but were later "co-opted" for a contemporary purpose (Gould

& Vrba 1982).

Although "exaptation" may not be an appropriate

description for the development of IgE, we cannot eliminate

it as a possibility. Perhaps IgE did originate for a

similar purpose as the other immunoglobulins, but was also

efficient in functioning as an anti-helminthic mechanism.

By adopting this approach, we may include in our analysis of

human evolution other traits which have previously received

little attention, because "complete" adaptation was not the

end result.

Influencing Biological and Sociocultural Factors

People who live in developing regions of the world do

not exhibit a high prevalence of allergic disease (Godfrey

1975; Warrell et al. 1975) presumably because of functioning

IgG blocking-antibody. This is probably because exposure to

parasites is chronic due to their traditional modes of

production and because they have limited sources of clean

water and sanitary provisions. When they shift to a more

developed, urbanized environment the incidence of allergy

becomes comparable to incidences in developed countries.

This is, however, an over-simplification of the allergy

model. A number of biological and behavioral factors

contribute to allergic disease, as will be demonstrated


Genetic. MHC-related factor

The development of allergic disease in an individual is

multi-causal. Perhaps one of the most predictive factors

for the development of allergy is genetic predisposition;

some individuals are genetically predisposed to produce more

IgE than others (Willcox & Marsh 1978; Marsh et al. 1980b;

Gerrard 1985). The phenotype of elevated serum IgE is at

least partly dictated by a genotype controlled by the MHC.

The MHC is a region on chromosome 6 in humans that contains

a number of gene loci. A particularly important complex is

composed of the four human lymphocyte antigen (HLA) loci,

designated as HLA-A, -B, -C and -D. Each locus contains a

number of different alleles, or HLA antigens, resulting in a

high degree of polymorphism. A set of HLA genes (that

includes four HLA determinants) constitutes a "haplotype".

Certain HLA antigens--or haplotypes--have been correlated

with immune responsiveness to aeroallergens (Menser et al.

1975; Blumenthal et al. 1980; Marsh et al. 1980b; Marsh et

al. 1981; Brostoff & Hall 1989). For example, HLA-Dw2 is

highly associated (92%) with allergy to the short ragweed

allergen Ra5, in contrast to a poor association (22%) for

those individuals with HLA-B7 (Marsh et al. 1981).

Genetic. non-MHC-related factor

High IgE levels are also dictated by a regulator gene

that is not linked to the MHC; total IgE is partially

regulated by an autosomal gene in which the genotype that

controls for high serum IgE is recessive (rr) and low total

IgE is dominant (Rr or RR) (Marsh et al. 1974; Gerrard et

al. 1978; Willcox & Marsh 1978; Marsh et al. 1980b; Rao et

al. 1980). It is postulated that the dominant allele (R)

functions by limiting the number of IgE antibodies that

clonally expand in response to an allergen, resulting in low

total serum IgE (Willcox & Marsh 1978). While the MHC class

of immune response (Ir) genes are antigen-specific, this

second genetic mechanism involving the regulator allele "R"

is nonantigen-specific (Willcox & Marsh 1978; Marsh et al.


Familial factors

Predisposition to allergy is associated with a positive

family history of allergy (Gerrard et al. 1976; Gerrard et

al. 1978; Marsh et al. 1980b), although the familial

environment might be a more important factor; members of a

family living in the same household and sharing the same

behaviors might all be at a similar risk of developing

allergies. Nevertheless, results from studies involving

twins conclude that serum IgE is genetically determined and

monozygous twins express more similar levels of IgE than

dizygous twins (Bazaral et al. 1974).

Ethnic factors

Prevalence of clinical allergy reportedly has been

higher in non-European descendants living in developed

countries, including the following: Chinese Americans (Worth

1962); West Indian Blacks in England (Davis et al. 1961,

Pearson 1973); American-born Filipinos (Orgel et al. 1974);

and Polynesians in New Zealand (Waite 1980). Similarly,

Iraqi, Iranian, and Yemini immigrants in Israel had higher

rates of asthma than other ethnic groups (Asch et al. 1973).

In contrast, some races exhibit a lower prevalence of

allergy than European-descendants; Herxheimer & Schaefer

(1974) reported that the incidence of asthma among Canadian

Eskimos was extremely rare. Asthma was also rare for North

American Indians despite their unusually high levels of

serum IgE (Gerrard 1985). It has been suggested that the

low incidence of asthma among Canadian Eskimos and North

American Indians may be related to the low frequency of HLA-

A8 haplotype in that population; white children with HLA-A8

have a particularly high incidence of asthma (Menser et al.

1975). However, Gerrard (1985) noted that serum IgE levels

were high in the Indian population; he hypothesized that

inadequate medical services meant that the North American

Indian was forced to rely more heavily on his/her own immune

system, possibly resulting in higher IgE levels.

Behavioral and nutritional factors

Similarly to other chronic diseases of modernization,

genetics alone cannot explain the rapid increase in the

prevalence of allergy. Just as the environment probably was

important for the development of IgE--that is, exposure to

an environment plagued with intestinal helminths--so the

environment must be important for the development of another

IgE-related manifestation: Human allergy. After all,

elevated IgE cannot manifest itself as a hypersensitive

reaction unless there is some aggravating allergen present

to elicit the response.

One factor characteristic of modernization and

important in the development of allergy is both biological

and behavioral, and that is breastfeeding. Breastfeeding is

believed to decrease the likelihood of developing allergy

indirectly by minimizing infection. Breast milk contains

factors that promote the maturation of the intestinal tract

of the newborn and also provides secretary antibodies to

assist in immunity at the intestinal surface (Ogra & Ogra

1978). Also, breast milk contains fewer foreign proteins

than bottled milk and it is believed that both the

introduction of foreign antigens (Johansson & Bennich 1985)

and infection during early childhood (Marsh et al. 1981) may

initiate allergic disease.

It is often difficult to differentiate between true.

food allergies and genetically-based food intolerances (see

Lieberman & Barnes 1990). A number of the food intolerances

(e.g., celiac disease, G6PD-deficiency, and lactose

intolerance) have been managed by culture-specific

proscriptions and food preparation practices in regions

where they are most commonly distributed. However, when

individuals with food intolerant predispositions are exposed

to new foodways, either voluntarily or involuntarily,

adverse allergic-like reactions often occur. Newly

introduced sources of dietary proteins--particularly the oil

seeds and yeasts--as well as the chemical modification of

foods for large-scale commercialization have potentially

antigenic effects on persons predisposed to allergic disease

(Metcalfe et al. 1988). A well-publicized phenomenon is the

"Chinese Restaurant Syndrome," which is an adverse food

reaction to monosodium glutamate (MSG) (Man-Kwok 1968; Allen

& Baker 1988).

Miscellaneous behavioral factors

Other behaviors associated with modernization
contribute to the development of allergy. Cigarette smoking

is associated with elevated serum IgE (Gerrard 1985; Burrows

1989). Psychological stress has been implicated in at least

the exacerbation of allergy and possibly as a predisposing

factor (Graham 1967; Glazer 1969; Smith 1978). Stress is a

well-known insult in populations undergoing transitions from

rural to urban settings, immigration, and an increase in

complexity of modes of production (McElroy & Townsend 1985;

Goodman et al. 1988). The disproportionately high levels of

allergy among Iraqi and Yemeni immigrants into Israel was

partially attributed to psychological stress (Glazer 1969).

Population behavior and the macroenvironment

As the environment in which we live becomes more

complex and we become more mobile, we are increasingly

exposing ourselves to greater numbers of foreign proteins

that may function as allergens. Immunologists conducting

studies in tropical regions have found that prevalence of

allergy tends to be much lower in populations where

traditional means of living continue compared to prevalence

in more developed populations (Anderson 1974; Godfrey 1975;

VanNiekerk et al. 1979; Dowse et al. 1985). Specifically,

urban populations overwhelmingly present with higher rates

of allergy than rural populations, and this discrepancy is

most notable in developing countries; examples include the

Gambia (Godfrey 1975), Nigeria (Warrell et al. 1975), and

the Punjab (Corruccini & Kaul 1983). With the process of

urbanization and modernization, populations are subjected to

new foods, ingested chemicals, and synthetic materials, as

well as chemicals and pollutants emitted into the

environment from factories and transportation vehicles.

Urban pollutants such as sulfur dioxide, sulfuric acid,

carbon monoxide, and particulate matter serve as irritants

to asthmatics (Lopez & Salvaggio 1978; Smith 1978; Hackney &

Linn 1985; Weiss & Speizer 1985).

Population behavior and the microenvironment

The role of the domestic environment--or rather,

"microenvironment"--was discussed at the outset of this

chapter. Specifically, some of what is known regarding

allergens and the microenvironment is related to the

presence of arthropods. Also important is the production of

molds (Burr et al. 1985; Brunedreef et al. 1989; Platt et

al. 1989; Dales et al. 1991) and fungi (Arundel et al. 1986;

May et al. 1986) secondary to excessive moisture levels in

the home. Inadequate ventilatory practices in housing

construction contribute to elevated relative humidity levels

(Arundel et al. 1986).

Conclusions and the Allergy Disease Model

Figure 1-1 illustrates the cultural and biological

factors sufficient--but not necessary--for the development

of allergic disease. The biological and cultural factors

are a result of interactions between the host population and

the environment; both factors influence each other, from the

host, and to the host. These interactions create elements

in the environment that subsequently produce allergenic

stimuli, or "insults", which in turn affect the host.

Depending on inherent features of the host, they may or may

not produce disease.

In summary, by reviewing many of the known factors for

the development of the hypersensitivity response, it is

evident that allergy is a disease characterized by human

modernization and urbanization. By using the construct of

the epidemiological transition we can predict the

development of allergy as a significant form of morbidity

due to changes which typically occur during the transition

of traditional living toward modern living. It is apparent

that populations in the middle of this transitional period

will experience an epidemic-like form of the disease due to

the sudden exposure to numerous allergens. Many of those

individuals may be more likely to suffer from allergy

because of their genetic make-up, although this is not a

limiting factor. Evaluation of disease patterns outside of

the Western world and into regions where the disease-causing

relationship begins will provide medical researchers with

new insights into the evolution of many human diseases--in

this case, allergy.



1. "Arboallergen" is a term used to denote any arthropod-
borne allergen. It is based on the hypothesis that a number
of specific epitopes are found in different arthropods, and
possibly occur in proteins associated with the production of
chitin and molting and/or in common digestive enzyme systems
(Brenner et al. 1991). The resulting clinical picture is
cross-reactivity (hypersensitivity) to different arthropods.
















The Construct of Allergy and Modernization:
A Review of the Objectives

In Chapter One, allergy was defined as part of the

epidemiological transition of disease and both genetic and

behavioral factors were identified that are known to

contribute to the rising incidence of allergy on a worldwide

scale. It was suggested that a closer look at a population

in the midst of infrastructural transition may elucidate

factors that are playing a significant role in the incidence

of allergy in both societal settings--developed and

developing. Certainly, identifying etiological factors of

disease is more complicated in a developed society, after

the epidemiological transition has occurred, in that the

population as a whole has already achieved a certain degree

of adaptation in its effort to rally against the disease,

and it is therefore difficult to distinguish between causal

versus spurious co-variation of suspected independent


The overall objective of the study--identifying a

relationship between modernization of the domestic

environment and allergy to household pests--was presented in

Chapter One. The dependent variable in the study is allergy

to a given household pest at the individual level, a


biological measurement to be presented categorically; a

person is either reactive or nonreactive to a specific pest

allergen. To minimize variation in selection criteria and

symptomatology associated with atopy, only asthmatics were

chosen for the study. The independent variables in the

study are structural features of the domestic environment

and sociodemographic features of the household that may

contribute to the nutritional and reproductive needs of

various household pests, thus supporting a substantial

population of the pests. The presence or absence of these

various features will be correlated with the presence or

absence of a given pest allergy to support a causal

relationship. It is recognized that categorization of the

independent variables as a feature of modernization is

primarily a subjective process; however, an anthropological

presentation of the evolution of the Barbadian domestic

environment will elucidate the dynamic features of housing

in Barbados and illustrate the fact that Barbadians are

achieving a level of development similar to that of

industrialized countries such as the U.S..

This chapter presents the general materials and methods

relied upon to conduct the study. Following an ethno-

historical review of the development of land distribution

and housing in Barbadian society (Chapter Three)--which

ultimately affects the development of the domestic

environment and thus the presence of certain pests--specific

sociodemographic and housing variables will be identified

and analyzed in Chapter Four. Chapter Five presents

findings from the entomological survey as well as stated

behaviors, perceptions and attitudes about household pests.

Chapter Six is a presentation of the biomedical testing of

allergies to various household pests, including results and

discussion, and Chapter Seven concludes this dissertation

with a summary of findings and implications for future


The Research Site
Barbados was chosen as the research site for a number

of reasons. An important sociodemographic feature was the

fact that Barbados is the most heavily populated country in

the Western Hemisphere, at over 1500 persons per square

mile. Household infestations are a significant problem in

densely populated settings (Harwood & James 1979) and also

in the tropics (Marchand 1966; Pearson & Cunnington 1973;

Dowse et al. 1986; Lan et al. 1988). Because the Barbadian

population is almost equally divided between rural and urban

residence and because the island is so small--and therefore

easily accessible with clearly defined boundaries--Barbados

offered a unique opportunity to contrast the incidence of

allergy to household pests with crowded urban and rural

populations. Another sociodemographic feature was the

dynamic status of Barbadian housing. The level of

infrastructural development in Barbados and documented

health indicators assured that the epidemiological

transition from acute and infectious disease to chronic

disease had in fact taken place in Barbados, thus implying

the likelihood of a significant incidence of allergy in

general and asthma in particular.

The Rising Incidence of Asthma in Barbados

Despite the absence of any recent study on the status

of asthma in Barbados, it is unanimous among most

Barbadians--health care professionals as well as lay

persons--that asthma is on the increase. While exact

figures for the incidence of asthma is not known, PAHO

reports that the disease category

"bronchitis/emphysema/asthma" accounted for a crude death

rate (per 100,000 population) of 8.9 in 1988, or 23 persons

in Barbados; this figure correlates highly to the U.S. crude

death rate of 9.2 for the same category and year (1990).

The number of asthmatic attendances at the Accident and

Emergency (A&E) Department at Queen Elizabeth Hospital (QEH)

doubled within a 10-year period, from 3,503 visits in 1980

to 7,137 visits in 1990 (Figure 2-1). Annually, asthmatic

attacks account for more than 13% of all visits to the A&E

Department, averaging about 20 patients a day (Naidu 1988,

1990), and make up 12.5% of the Emergency Ambulance Service

(EAS) calls (Naidu 1992). A slight decrease in the numbers

of visits during 1989 and 1990, at 7,267 and 7,137,

respectively, suggested that the dispensing of free asthma

medications and the increase in public awareness and

educational programs were possibly stabilizing the incidence

of asthmatic attacks (McCarthy 1991a). But a record 7,808

visits in 1991 indicates a continuation in the trend of

rising incidence. Nearly half of all asthma-related visits

are children under the age of 15 (McCarthy 1991b).

The incidence of asthmatic attacks has a definite

seasonal pattern in Barbados. October is the month with the

greatest number of visits; Figure 2-2 illustrates the

dramatic increase from 1974 to 1991. There were an

unprecedented 936 visits in 1988, 899 in 1990, and a record

937 visits in 1991 (Naidu', personal communication). Also

in 1990, two of the three asthma-related deaths occurred in

October (McCarthy 1991c). This corresponds to the rainfall

patterns in Barbados, in which the average monthly rainfall

peaks in August and October (Depradine et al. 1984).

Figures 2-3 and 2-4 illustrate the relationship between

asthma visits to the A&E department and monthly rainfall for

1990 and 1991, respectively. Note the unusually large

amounts of rainfall for the month of October--and the

exceptional amount of rainfall for November, 1991--

correlated with the highest numbers of asthmatic visits on a

monthly basis.

/ 1e o

9repuavjO "ONfI










V *
w 0


0 -5
I-0 QM

g ..
# (0

*e-4i 0
*4 aO
A to

S---------* ii

-- ------ "861 .

------------ ---* 9861

"-""" -- -I-861

.. .......- .... .....6
~~~--------- ------6

-~ : : -- --- : ---- 861

-- ---. ,---. 186 1

- ----------^ 0861

----- --- -. 9L61

L... 6T

91o 61


s *uupum-pV jo Joqum N





V :0


V 0-ri


tv to

0 -H 0
04 FA

4- 0 *d

to M
N N z
I -H


0)4) :3
'U-4 :

I -'-4

0 0

Asthamtic Attendances

0 00 00001000 0
oo m or^ t- Co D 00 m Rt R

saqouI ul ffJ" PI






V 0










Asthmatic Attendances

o o

"t 0 t0 C
Clq N "-I "-I

SaL~uJ UTuju-m-













t U

00 4 0

The significance of the rising incidence of asthma in

Barbados is reflected in the facilities that have been

created in recent years to cope specifically with the

asthmatic patient. The current Minister of Health, Mr.

Branford Taitt, became particularly interested in asthma and

health care, due in part to the number of deaths from

asthmatic attacks and to the apparently overall increasing

incidence of asthma. A policy was instituted in which

polyclinics would have nebulizers, peak flow meters, and

oxygen available at all times, and asthma drugs would be

available in all polyclinic dispensaries. In 1990 the A&E

Department was remodeled and included a new "asthma bay"

designed especially for asthmatic clients. The bay is

situated directly in front of the nursing station, thus

allowing full-time observation. An important consideration

in understanding the increasing number of hospital visits

due to asthmatic attacks is a greater reliance on the A&E

Department than in previous years. Naidu suggests that this

trend may represent an increasing dependence by the public

on the A&E Department as a primary care center for the

treatment of asthmatic attacks (1990). It has also been

suggested that the decrease in reliance on home remedies and

increasing preference for A&E medical treatment is

responsible in part for the increase in attendances

(Ferdinand2, personal communication; Naidu, personal



Unfortunately, there are few physicians in Barbados who

specialize in asthma disease. One allergist conducts

allergy testing and desensitization therapy in his private

clinic, but the cost of his services are prohibitive to the

average Barbadian. A few physicians, both in the private

and public sector, who specialize in internal or respiratory

medicine, have earned a reputation as "asthma specialists"

and are preferred by asthmatics and parents of asthmatics

over polyclinic physicians. In 1990, medications for the

treatment of asthma were made free to all asthmatics through

the Barbados Drug Service. For the period between October

1989 and December 1990, 5,310 asthma therapy prescriptions

were filled at the QEH pharmacy and 7,892 were filled at

other government pharmacies, for a total of 13,202

prescriptions (Prescod3, personal communication). Money

spent on asthma medications actually decreased by 4.5% from

BDS4 $261,039 (note: all monetary figures in this thesis

will be presented in Barbados dollars, except when indicated

otherwise) between July 1988 to June 1989, to $249,435

between July 1989 to June 1990; this factor has been

attributed in part to the efficiency of the Drug Service

program (Prescod, personal communication).

In June, 1989, the Asthmatic Association of Barbados

was founded, following suggestions for the organization

during a "national asthma week" seminar provided by the

Barbados Drug Service. As of 1991 the organization boasted

some 150 members, most of whom were adult asthmatics or

parents of asthmatic children, although the meetings are

open to any interested person. The association meets once a

month and the primary objective is to disseminate

information on the control and management of asthma, usually

via a guest speaker5.

Infrastructure: The Developed Nature of a Developing

In 1991 the United Nations revised the human

development index and declared Barbados the leading

"developing" country among all developing nations (United

Nations 1991), a step that Barbadians fear will eventually

reduce their eligibility for international aid. The

classification is supported in part by an estimated per

capital income of US $6,020, higher than the Republic of

Korea or oil-rich Venezuela (United Nations 1991). Thus,

according to the U.N., Barbados ranks ahead of its neighbors

in the Eastern Caribbean in human development; Table 2-1

illustrates the disparity of GNP (gross national product)

per capital within the Eastern Caribbean nations.

Barbados health indicators demonstrate a quality of

life more comparable to industrialized countries than

developing countries (Table 2-2). Great strides have been

made in Barbadian health care; between 1920 and 1922, life

expectancy was only 31.9 years for women and 28.5 years for

Table 2-1 GNP per capital (1980), in U.S. dollars in
the Eastern Caribbean.

Country GNP per capital

Barbados 6,010
Antigua & Barbuda 3,690
Trinidad & Tobago 3,350
St. Kitts & Nevis 2,630
Grenada 1,720
Dominica 1,680
St. Lucia 1,540
St. Vincent 1,200
Jamaica 1,070

Source: United Nations 1991.

men (Dann 1984). This is in stark contrast to the average

of 75 years as of 1990 (United Nations 1991). In a "quality

of life" survey, sociologist Graham Dann found that "health"

received the highest score of the seven life satisfaction

domains (1984), indicating that Barbadians are content with

the status of their health and health care delivery. The

leading causes of death in Barbados are similar to those of

industrialized nations, and include: heart disease,

malignant neoplasms, cerebrovascular disease, diabetes

mellitus, and "other" circulatory system diseases (PAHO


Education in Barbados ranks as the leading priority in

government spending, at over one fifth of the national

budget, or, about $1 per person per day (Dann 1984). Table

2-3 compares educational indicators in Barbados with other

Eastern Caribbean countries. As illustrated, there is a 99%

literacy rate and education is compulsory for 11 years. It

has been argued that Barbadians are perhaps over-educated

for the jobs available, as unemployment for new graduates is

higher than in any other age group (Dann 1984). Recently

there have been efforts by the Ministry to design a

curriculum that would prepare students for the three

mainstays in Barbadian economy--tourism, manufacturing, and

agriculture, but this move has been sharply criticized by

certain nationalists as "encouraging a 'mental attitude of

subservience'" (Goddard 1991, p.6).

Barbados was controlled by Britain for some 350 years,

and at the time of Independence (1966) the British legal and

parliamentary system was adopted. Barbados is classified as

an independent, liberal-democratic state with competitive

parties, has the highest rating for political and civil

rights (according to the Freedom House Index) and a strong

emphasis on economic and social reforms (Stone 1985). There

is a network of approximately 860 miles of paved roads and

public transportation is available in all of the 11

parishes. PAHO (1988) reported 100% coverage of drinking

water supplies and sewerage or excreta disposal services in

Barbados, which no doubt contributes to the overall good

health of the nation. (See Chapter Three for a full

description of Barbadian economy).

The Health Care Delivery System

Barbadians enjoy an efficient socialized health care

system, reflected by the health indicators previously

illustrated. Government health expenditure per person in

1987 was US $230.58 and the percentage of total government

expenditure on health was 13.13 (PAHO 1990). The primary

resource centers for receiving health care in Barbados are

QEH6, including the hospital and polyclinics; the

distribution of these services is depicted in Figure 2-5.

The concept of polyclinics and public health care came

into existence in the mid-1970s. Initially the clinics were

serviced by General Practice (GP) doctors who maintained

both government and private offices and visited the clinics

several times a week. At that time, only patients who

passed a "means" test by the welfare department could visit

the free clinics or the hospital. In 1985 the government

began employing GP's full-time and regular GP sessions

commenced in the polyclinics. The "means" test was

eliminated and services became available to all residents.

While the exact number of persons attending private

physicians is not known, 135,810 attendances were recorded

in 1990 to the eight public polyclinics and five outstations

(see Table 2-4) (Sergeant7, personal communication).

According to R. Naidu, Director of the A&E Department at

QEH, the introduction of the polyclinic system has reduced

the number of visits to the A&E Department, except in the

parish of St. Michael, where there is still a

disproportionately high number of attendances when compared

to the percent of the population living in that parish

(Naidu 1988). The Barbados Drug Service provides free

prescription medications to all persons under the age of 16

years and those over 65 years and to any patient needing

medications for venereal disease, cancer, hypertension,

diabetes, glaucoma, and, as of 1990, epilepsy and asthma.

Dispensaries are available at each of the polyclinics.

4H .4.O.4
4JU oa-4

C 0 i00

0 00
- 4 V -H
41 4.) CO
I o co
rd m co
0 0 0 COW
P4 024. fa
04J UC
0 EU 0 -..

.,-I 0
4 4) -m
0 W, -- CO
- .0 4)0 4) to
3 o) co

0 0 -
4H 4 o
to M r. U CO
34 ) r-4 U I
0V U (0 l 0O
0 U 4) 1. 4n

0 -0
04 (C A
4 0 E cn--I
c XUlC 0
1^ D(0 **1 r-



CN OD 0 .
CVM H -I r)













I *


I l
I 8


I 9

01< c

'0 0
0O -

* I-
. M

0 0
01 4-I

go AS 1
k 9 t 0 *(- U 0
E-4 Q m 0 t-

*M4 C *0 00 0
$4 .a < C *U .
CO *M ,Q b U
*Ml E-* MM 5i P
k C4O4
Psa dmQMO~ [


D a O \In

-I 01

* *
" 0 CM

S Co

) I




. >

I C f 0
> E0 $
1 >i
40 1 0
ca 010
C/, C3IU3




0 0 0
o W-r 0


) co
- -4

S 00
0 34.)
+1 0
C UC03

o r-4 ), 4) co
04 V A 00
I M apco
O 0 3 1C-4

0 a -4 o4

= .0 C e-4 0 OI
0 4 0 0 *4-I W

-1 U k V co
4)a 4 4I -4

-4 0) U)
3 4.) 4V co
V -'-4 0 d oC
S 4 r-4 4 -H






I (
I (

I (


I l
I u


I s


1H 0 H-H r- I

YOI .I 44W
U) r1 riI r0 U) U) W
0 I

%Dl a r-I V w fl H

0 I l *
0 to FA
tr to *<
U0 p z

4 -

"A 41 A 0 0 *
E-i ( Q hC 0 ) M* M ) 1I
^ C O (0 ^ 4- +

+ Polyclinic

t. Peter St. | Accident & Emergency Dept.
SAndrew and outpatient clinics, QEH

James St. Joseph

St. Thomas
St. John
St. George
S St.
Michael St. Philip

Christ Church

Figure 2-5. Distribution of Outpatient Health Care
Facilities in Barbados.

II I 63
II IooE )O I
II ve-Ii inI
II > 0 1 I
II vCI l ol
II nIr m~0woHwwoo i mig

II I I 10
II I I -,4
II W I I 14 .

IV 0 91, O4 0 1 0 I
II i I 1 0

II 0 I 1

II Cl I l-e

1 4)
1 0
II ) >10 I I
SII .Ci N r4n NOlI
d 4),100oi N omNo OOvN %0o 0 -IO
II 0 > 04. 0 o o0 -C N N 1 0
0 I lI

II I 0
II .I I i

II I 1
S II 0 1 I g

to1 V to
SII O .1 n n I 01

SI tI1-,-4 V CO V qr IM I -4 I
V II .t I .o. M- OM IN I
II E-4 p a V Nr N n o 01 v r- in in mIi
*11 04 1 00 -CO N -A-4 -to V)H ii A

I I I 1I

II I I 1


C- II 01 Of O
0 II I 10
0 II M I C I 1W
CIo 4-- I) N 1L 4in f i CI I

4:1 II ,-H --iin, T>1 0 1 fa.,- o -In i 0 m 1:
SII4 I -, -I4 =0 0 fI O x 1-I- I 1 1 .
H 0)=1 C Ai OHN -p'0 -c) -0-I1 14)

c II I I )
411 CI w to rI v.- 1 >i

AI XI l H W -i H C-i -4 1 10
4)11 IUH MU :WW +woI

I I* *H I C

*-4l II I I .

II I p 0 c

O4 S 01 I C
S) 0 -H II10
U1 13 >iH C WI I
0Ill I-0V U -'-I 18 0 WI I
I II CH 0*0<1 W I 0O
011 -I UVfl00l)aMOH lr41 01

Sampling and Methodology

The Survey Schedule

The asthma study was conducted in Barbados from January

to December, 1991. The project was divided into 2 phases.

The principle objectives of Phase I were the following: (1)

to qualify the household pest species in the major

topographical regions of the island; and (2) to conduct an

ethnographic survey, primarily using the techniques of

participant observation and informal interviewing, for the

purpose of designing an appropriate questionnaire for the

study sample in Phase II. These separate operations were

conducted simultaneously; while the trapping was taking

place in one region, the ethnographic data were collected in

the selected households. Three to four weeks were spent at

each site during the dry season and two weeks were spent at

the sites during the rainy season.

Topographical considerations

Unlike its volcanic neighbors, Barbados is composed of

a coral limestone formation, resulting in a relatively flat

but terraced landscape, with a series of deep gullies as

well as vertical cliffs, the remnants of old coral reefs.

The coral limestone base contributes to a good drainage

system and a fluctuating but reasonable supply of water,

depending on the region, yet negates the presence of rivers

or streams. Although there were at one time a number of

coastal swamps on the leeward coast, the only remaining

swamp--and the only true body of water on the island at

present--is the Graeme Hall Swamp, situated in suburban

Christ Church.

The Town and Country Development Planning Office (1988)

has divided Barbados into nine topographical areas (Figure

2-6). In the northern cap is the flat St. Lucy Plain (1).

Below the First High Cliff area (2) lies almost the entire

western beach coast, 20-25 meters (66-82 feet) above sea

level. Below the Second Cliff area (3) the region is

divided into three sub-regions of terraces. About 35 meters

(116 feet) above sea level is St. George's Valley (4), a

synclinal structure and relatively small component of the

island, and to the east of that lies St. Philip Plain (5), a

very flat and poorly drained region. To the north and in

the center of the island is the Upland Plateau (6), with

dramatic variations of gullies and low terraces to

elevations from 130-330 meters (430-1090 feet). An

anticline of over 130 meters (430 feet) makes up the

southern Christ Church Ridge Unit (7). Finally, there is

the Scotland District (8 and 9) in the northeast, the most

distinctive of the regions, due to the heavy erosion and

subsequent exposure of underlying geological formations

consisting of shales, sands, silty clays, and marl (see Town

& Country Development Planning Office 1988).

Figure 2-6. Geophysical Subdivisions of Barbados. Numbers
correspond to the following regions: (1) St. Lucy Plain,
the flat, northern cap of the island; (2) First High Cliff
area, below which lies the western beach coast, 20-25 meters
(66-82 feet) above sea level; (3) Second Cliff area, divided
into three sub-regions of terraces; (4) St. George's
Valley, 35 meters (116 feet) above sea level; (5) St. Philip
Plain; (6) Upland Plateau, with elevations from 130-330
meters (430-1090 feet); (7) Southern Christ Church Ridge
Unit, an anticline of over 130 meters (430 feet); (8) & (9)
Scotland District. Figure adapted from map obtained with
permission by Town & Country Development Planning Office

There is significant variation in annual and seasonal

rainfall and, because of the terraced structure of the

island, rainfall varies considerably from region to region

(Figure 2-7). The Barbados Society of Technologists in

Agriculture compiled a list of monthly and total rainfall

figures in Barbados from 1847 to 1983 and found an annual

low of 39 inches in 1947 to a high of 91.5 inches in 1901

and an average annual rainfall of 59.68 inches (1984).

Seasonally, however, rainfall averages a low of 1.90 inches

in March and over seven inches per month from August to

November, during the "rainy season" (Barbados Society of

Technologists in Agriculture 1984). Figure 2-7 illustrates

the variation of rainfall within the different regions of

Barbados, showing the greatest amount of rain falling in the

central portions of the island (an average of 80 inches) and

the lowest amount in the southern Christ Ridge region and

St. Philip (an average of 45 inches) (Plantations Trading

Co. Ltd. 1989). Areas with the lowest rainfall correlate

with the three areas with the least fertile soil, the west,

south, and southeast coasts, and these are also the regions

that are the most densely populated (Town & Country

Development Planning Office 1988).

It is apparent from the above discussion that, although

Barbados is divided into 11 parishes, parish boundaries do

not necessarily represent the topographical boundaries. The

Leeward side of the island--or West and South coast--is the

most heavily populated and comprises the four major urban

centers of Barbados: Speightstown, Holetown, Bridgetown,

and Oistins. Within the metropolitan boundaries of Greater

Bridgetown live 45% of the national population and the

majority of governmental operations and industry. The area

between Bridgetown and Speightstown is predominantly devoted

to tourism, as resorts line the length of the coastline.

Speightstown is a semi-urban fishing town, populated by

Barbadians rather than transient tourists. The region south

of Bridgetown toward Oistins is also a tourist region,

though this area is mixed with residential areas and

businesses and is the seat of the main fish market.

Recently there has been considerable urban tourist-related

growth in a southeasterly direction, encompassing the parish

of St. Philip. The major center in this region is Six

Roads. Although the eastern portion of Christ Church and

the parish of St. Philip are increasingly the focus of

growth for industry, the region is still largely

agricultural. The northeastern Scotland District is the

most sparsely-populated region on the island.

Figure 2-7. Annual total rainfall by region, in inches,
based on data from 1887-1986. Compiled by Barbados Society
of Technologists in Agriculture. Figure adapted from map
obtained with permission by Plantations Trading Company
Ltd., 1989.

Selected entomological and ethnographic research sites

After conferring with members of the Ministries of

Health and Agriculture, it was decided that five general

areas would be systematically selected in order to represent

the major topographical regions of Barbados. Villages

included the following:

1) Chapman Street Village in the center of
Bridgetown, representing the most densely-
populated, urban region;

2) Oistins Christ Church, representing a semi-
urban, coastal population;

3) Cotton Vale a rural tenantry, north of Six
Roads, representing the St. Philip Plain and
agricultural sector;

4) Chalky Mount another rural tenantry, but
situated in the northeast Scotland District of St.

5) Rose Hill a semi-urban village, below the
Second Cliff area, just south of the St. Lucy
Plain and north of Speightstown, St. Peter.

Once a village was selected, the houses were mapped,

numbered, and listed. From that list four homes were

randomly selected. The adult present at the time of the

visit was asked to participate in the study. If a

consenting adult was not home, two subsequent visits were

made in an effort to include the household in the study. If

after the third visit no adult was home, or if that adult

declined to participate, another house was randomly

selected. This continued until four households in each of

the five villages agreed to participate. Figure 2-8 depicts

the location of the five sites for Phase I.

Entomological samples were collected during both the

"dry" season (January-May) and the "wet" season (September-

December). The methods and materials are presented in

Chapter Five, after which the results and discussion of

trapping are addressed.

Participant Selection and Resources

In order to select the sample population for the

biomedical testing, the following persons were consulted:

the acting Chief Medical Officer, a Ministry of Health

epidemiologist, Medical Directors of six polyclinics, the

Director of the A&E Department at QEH, a Respiratory

Medicine Specialist at QEH, and a private Internal Medicine

physician in St. Michael. With the assistance of these

persons, three major health care resource centers were

selected for choosing the sample population and included the


1) five polyclinics and one outstation, the total
of which serviced patients from all regions of the

2) a private practice clinic, situated in the
urban district of Greater Bridgetown and servicing
patients from all of the 11 parishes;

3) and the A&E Department at QEH, servicing the
entire Barbados population.

Figure 2-9 illustrates the resource centers used for

selecting the sample population. Notice that two of the

polyclinics selected were located in the parish of St.

Michael to account for the large proportion of the

population living in that parish, and two of the other

polyclinics--Sir Randall Phillip and Maurice Byer--were

located in the urban "belt", in the major towns of

Speightstown and Oistins, respectively. Maurice Byer

services patients from the northern region of the west

coast, north of Holetown, the northern cap of St. Lucy, and

the northeastern Scotland District. The remaining rural

residents in the southern Scotland District region are

serviced in part by the Gall Hill outstation in St. John.

The private physician's clientele were mixed; many were from

the lower socioeconomic strata, who preferred to use a

private physician rather than a polyclinic and others were

middle- and upper-income clients, many of whom lived in the

suburban terraces and heights8.

Three-hundred and fifty children between the ages of

five and 18 years were systematically selected from the

health care centers. Each asthmatic patient attending one

of the selected clinics was asked to participate until a

total of 175 asthmatics was obtained; in the event that the

parent/guardian agreed to participate, the next non-

asthmatic child attending the clinic that matched the

asthmatic by age and sex was selected, until 175 controls

were obtained. Because of the systematic selection process,

if the patient to be seen was accompanied by siblings or

relatives, asthmatic or non-asthmatic, the other children

were asked to participate as well. The only exclusion

criteria was that the child must either be an asthmatic or

non-asthmatic; other types of atopics were not included

(e.g., hay fever). Selection of asthmatics was based on a

positive history and physical examination and criteria

included previous attacks, diagnoses and treatment for

asthma. The history and physical are the primary means for

diagnosis of asthma in Barbados; diagnostic tests such as

serum titers, bronchial challenge tests, and skin tests are

not practiced routinely, largely due to prohibitive costs.

A "selection criteria" form was designed, approved by the

acting Chief Medical Officer, and completed for each child

(asthmatic and control) to confirm or eliminate a diagnosis

of asthma.

At the time of selection the parent/guardian of the

child read and signed a written informed consent and was

assigned an appointment for the allergy testing at one of

the six chosen polyclinics. The parent/guardian was asked

to choose the polyclinic that was closest to the child's

home and instructed to withhold antihistamines, ephedrine,

or asthma medications 24 hours prior to the testing. In the

event that medications had to be administered, the

appointment was rescheduled. Parents were reminded of the

appointment by telephone 24-48 hours before the test. If

that household did not have a telephone, the nearest

relative or neighbor with a telephone was contacted.

Clinical Testing and the Interview Schedule

The allergy test included skin testing (scratch and

intradermal) to various household pests (note: details of

the skin test procedure are described in Chapter Six).

Controls did not undergo the skin test due to the difficulty

in recruiting non-asthmatic participants for this aspect of

the testing; the skin test involved a battery of 11

extracts, requiring a minimum of 11 scratches and

potentially 11 injections. Receiving no personal benefit

from undergoing this procedure, parents and non-asthmatic

children were unlikely to volunteer as controls and it was

deemed unethical to subject them to the procedure. To

assure validity of the test within the asthmatic population,

a negative control was used in both the scratch and

intradermal test. Exclusion of non-asthmatics from the skin

testing procedure was further substantiated by the overall

objective of the study, which was to determine which

features of the domestic environment were contributing to

the likelihood of asthmatics developing allergies to certain

pest species.

Rose HII
St. Peter Andr w
0 halky Mount
JQ s St. Joseph

St. Thomas
St. John

St. St. George 0 Cotton ale
Mic ael
o St. Ph'ilip
Chap an
Street Christ Church


Figure 2-8. Location of selected communities for the
entomological and ethnographic surveys.

Private practice

Andrew \ l Accident & Emergency Dept.,
St. Joseph
St. Thomas\ E
St. John

+ St. St. George
Michael St. Philip

Christ Church

Figure 2-9. Resource centers for the asthma study sample

The questionnaire, designed from ethnographic data

collected during Phase I, was pre-tested on a randomly

selected population (N=10) prior to operationalizing Phase

II and included informants who were parents of asthmatic

(N=5) and non-asthmatic children (N=5) from the five

villages in which Phase I was conducted, but not from the

same households. The questionnaire was revised and pre-

tested again on another randomly selected population (N=10),

equally divided between parents of asthmatic and non-

asthmatic children and also from the five villages.

Following the pre-testing phase, the final

questionnaire was divided into two sections so as to

minimize the risk of informant fatigue and annoyance during

the interview. The first section was administered in the

clinic prior to the allergy testing and serum collection.

An appointment was made in the clinic for a home visit so

that the second section of the questionnaire could be

administered. In the event that a home visit was not

possible due to time constraints, a telephone interview was


The questionnaire included three general topics:

sociodemographic information on the child's residence and

structural information on the dwelling; the informant's

perceptions, knowledge, and behavior associated with health

and illness in general and asthma in particular; and the

informant's perceptions, knowledge, and behavior associated

with household pests and pest-related disease.

Sub-Setting the Sample Population

As described, the overall objective of the study was to

determine the relationship between sociodemographic and

architectural independent variables with the dependent

variable, hypersensitivity to individual household pests.

Seven asthmatic children and eight controls were eliminated

from the study after the trials and interviews because

either their classification of "asthmatic" or "control" was

questionable, or a follow-up interview could not be

administered. The group of asthmatic children who underwent

skin tests to determine the pests to which they were

allergic comprised the "asthma population" (N=168).

Controls (N=167) were used to compare informant responses

regarding perceptions, knowledge, and behavior related to

health care and asthma. The combination of asthmatics and

controls made up the total "study population" (N=335).

To conduct bivariate analyses on the sociodemographic

and structural variables and to construct a socioeconomic

index to be included in these bivariate analyses, it was

necessary to reduce the total study population to the

household level. The rationale for this breakdown is the

fact that 72 informants had more than one child in the

study; therefore, the weight of that informant's response

would be greater than that of an informant with only one

child in the study, resulting in a "pseudoreplication" of

responses and misleading correlations.

Consequently, one child from each household was

selected randomly from the total population, regardless of

his/her classification as asthmatic or control, using the

Q&A database program (Symantec Corp. 1991), and data on that

child and his/her household was entered into a subset

referred to as the "household sample". The total household

sample size was 177 (after disqualification of cases with

insufficient or ambiguous data). Bivariate analyses on all

of the independent variables were first administered on this

sample population to identify correlations and selection of

the appropriate independent variables that would later be

analyzed for the asthma population.

Identifying and Classifying the Independent Variables

The word "modernization" connotes the up-grading of

standard of living at both the population and the household

level. Standard of living at the population level is

typically measured by socioeconomic indices such as per

capital income, quality of housing, and quality of life

factors, including health. In this respect, Barbados fares

well according to U.N. standards.

Standard of living at the household level may vary

tremendously within a population and can be used to

represent the degree of class disparity in a given society.

While Barbados enjoys a high standard of living compared to

other developing countries, in no way is Barbados a

"classless society", as there clearly exists a significant

degree of class consciousness and stratification (Dann 1984;

Potter 1983a,1986). For example, nearly a quarter of the

population lived below the poverty line between 1980 and

1988 (United Nations 1991). In order to classify

respondents according to socioeconomic status, an attempt

was made to create a socioeconomic index variable.

Designing a Socioeconomic Indicator

Graham Dann, a Barbadian sociologist, conducted a study

entitled "The Quality of Life in Barbados" (1984), in which

he addressed a number of socioeconomic and sociodemographic

issues in contemporary Barbados. He discussed the weakness

of most studies in the Caribbean prior to his own study, in

that the classification of socioeconomic status was usually

based on a single indicator, such as income or occupation,

thus failing to recognize "...that social class is multi-

faceted" (1984, p.30). In response, Dann constructed a six-

item index for measuring socioeconomic status among his

sample population, which included the following: (1) social

mobility (based on the improvement of residential location

and real income); (2) income group (reported monthly income,

grouped into quintiles); (3) occupation (5 categories, in

ascending levels of skill); (4) home ownership; (5)

amenities, (household possessions); and (6) education

(primary, secondary, and tertiary). Dann's index was

integrated into the asthma study questionnaire, but during

the analytical phase of the project, certain weaknesses were

realized regarding the application of this scale to the

asthma study.


Reported income as a measure of socioeconomic status in

the Caribbean is unreliable for several reasons. One

problem is the predominance of a working class, which is

quite different than that seen in industrialized countries

(Dann 1984). Particularly in Barbados, credence must be

given to the importance of remittance as income secondary to

large-scale emigration (Roberts 1955; Cumper 1959; Ebanks et

al. 1979). Other influencing factors in household income--

and often undermined--are contributions from the informal

sector, underemployment, and the seasonality of'jobs. For

example, Dann (1984) reported that, of the 6,900 7,000

laborers in the sugar industry, 2,500 3,500 were seasonal

crop workers.

Not surprisingly, and probably due to a combination of

these factors and others, failure to report combined family

income was relatively frequent in the asthma study

population; 17 informants (9.6%) gave no information

whatsoever when asked what the monthly income was for the

household. Twenty-eight informants (16%) stated that she/he

had "no idea" what the entire household earned on a monthly

basis and could only report her/his own or one or more