Risk factors associated with house infestation by the Chagas disease vector Triatoma pallidipennis...

Risk factors associated with house infestation by theChagas disease vector Triatoma pallidipennis inCuernavaca metropolitan area, Mexico

J . M. RAMSEY1 , A . L . ALVEAR1 , R . ORDONEZ1 , G . MUNOZ1 ,

A . GARCIA2 , R . LOPEZ2 and R. LEYVA3

1Center for Infectious Disease Research, 2Center for Population Health Research and 3Center for Health Care Systems

Research, National Institute for Public Health, Cuernavaca, Morelos, Mexico

Abstract. Chagas disease caused by infection with Trypanosoma cruzi Chagas(Kinetoplastida: Trypanosomatidae) is widespread in Mexico, transmitted byvarious triatomine bugs (Hemiptera: Reduviidae). The only domestic vector inCuernavaca (population 650 000) is Triatoma pallidipennis (Stahl) with T. cruziseroprevalence ranging from 1% to 9% in the resident human population. Wesurveyed possible risk factors for T. pallidipennis infestation at Cuernavaca (alti-tude 1200–2200m) on south-western slopes of the Sierra Madre Occidental. Thismetropolitan area (with five administrative counties) has rapid urbanization,forested environs and proliferation of ‘weekend housing’ for visitors from MexicoCity, 60 km to the north.To assess factors associated with T. pallidipennis infestation, we first stratified

Cuernavaca by altitude and by socio-economic status of population catchmentunits (PCUs). Within each PCU, one to three blocks were chosen for clustersampling (three houses/block) and information about Chagas disease was distri-buted. After obtaining signed consent from householders, representative houseswere routinely and opportunistically inspected for T. pallidipennis and surveyedfor demographic, economic, physical and other potential risk factors.Of the 1129 houses assessed, T. pallidipennis was found in 4.1% (range 3.0–6.8%

per county) and the T. cruzi infection rate was �50% in bugs. Rates of houseinfestation in poor PCUs were double those in higher socio-economic strata (oddsratio 2.12, confidence interval 1.03–4.3), with >4-fold greater crowding index ofT. pallidipennis. The bug density index was inversely correlated with PCU altitudeand socio-economic category (altitude of homes being associated with prosperity),while the bug colonization index (presence of nymphs indicating breeding)did not vary significantly across the PCU categories, but did vary according toaltitude.Multivariate regression analysis showed that the most significant risk factors

associated with T. pallidipennis infestation were lower altitude (linked with lowersocio-economic status), garden area >80m2, dogs at liberty to enter the house,occurrence of squirrels and opossums around the house, presence of pigs in thesurrounding area and having at least one of the adjacent lots empty (unconstructed).Householders who had received information about Chagas disease comprised 33%

from infested houses (14/42) but only 15% from non-infested houses (148/984). Hence,the awareness of Chagas disease was significantly associated with having a bug-free

Correspondence: Dr Janine Ramsey, CISEI/INSP, Av. Universidad 655, Cuernavaca, Morelos 62508, Mexico. Tel.: þ52 777 329 2363;

fax: þ52 777 3175485; e-mail: [email protected]

Medical and Veterinary Entomology (2005) 19, 219–228

# 2005 The Royal Entomological Society 219

house (P< 0.01). When shown specimens of T. pallidipennis, the proportions of house-holders who recognized them were 78% from infested houses but only 29% of thosewith uninfested houses. Given the low infestation rates and the high capacity of thepopulation to act appropriately once they have received information regarding thisdisease and its vector, relevant health education is expected to have a significant impacton triatomine control in this metropolitan area.

Key words. Triatoma pallidipennis, Trypanosoma cruzi, bug infestation, Chagasdisease, health education, house structure, risk factors, socio-economic stratification,triatomine control, urban infestation, Cuernavaca, Morelos, Mexico.

Introduction

Chagas disease is endemic throughout Mexico: perhaps

85% of the 100 million population are at risk of acquiring

Trypanosoma cruzi infection via at least eight species of

vector Triatominae (Ramsey et al., 2005), while 6 million

are at risk through blood transfusion (Guzman et al., 1998).

We estimate that 69 000 individuals/year contract T. cruzi in

Mexico, while nearly 1.8 million (�2% of the population)

are currently infected with the parasite. With its insidious

pathology, Chagas disease contributes to 25 500–63 000

deaths/year in Mexico, including infant mortality of �830/

year (Ramsey et al., 2003b, 2005). The Chagas disease bur-

den for Mexico currently includes >530 500 cases in the

chronic phase of the disease, which usually begins during

maturity (aged 35–55 years) affecting productivity, and may

involve costly supportive therapy for cardiomyopathy. For

example, one of the main hospitals (operated by PEMEX)

in Salina Cruz, Oaxaca, reports that 85% of cardiac insuf-

ficiency cases (16% of all cardiac consultations) are due to

Chagas disease, even though patients attending this hospital

originate from an area with relatively low transmission

(Moreno Lopez et al., 2001). The overall economic burden

of Chagas disease in Mexico is estimated at US$ 3160 mil-

lion annually; if all patients receive diagnosis and supportive

treatment the cost could reach US$ 126 million/year (Ram-

sey et al., 2003b, 2005).

In most Latin American countries it is the rural popula-

tions who are most at risk of domestic colonization with

triatomine bugs and consequent transmission of T. cruzi

(Schmunis, 1996). In Mexico, however, urban transmission

of T. cruzi seems also to be common. Although nationwide

information is not available, the State of Morelos in central

Mexico reveals the urbanization of vector infestation: three-

quarters of the State population reside in urban/suburban

communities and >77% of them are infested with Triatoma

(SSM, 2001). Cuernavaca, the capital of Morelos, consists

of five municipalities with a permanent population surpass-

ing 650 000 (INEGI, 1995). Seroprevalence of T. cruzi in the

resident population ranges between 1.2% and 9%, while the

state blood bank reports 0.8% of transfusion blood dona-

tions to be seropositive. Cuernavaca has been the weekend

refuge for many inhabitants of Mexico City since the 16th

Century when Cortez created Cuernavaca palace to escape

the cold and humid conditions of the Texcoco and Xochi-

milco plains now occupied by Mexico city. The present

study quantifies the vector infestation rate and its spatial

distribution at Cuernavaca, identifying risk factors asso-

ciated with T. pallidipennis, through house-based sampling

stratified by altitude and socio-economic ranking.

Materials and methods

Sample selection

This study covered the Cuernavaca metropolitan area,

administratively composed of five counties (Table 1) with

communities divided into population catchment units

(PCUs): Cuernavaca county has 110 PCUs in one commu-

nity; Emiliano Zapata county has 24 PCUs grouped as three

communities; Jiutepec has 81 PCUs in four communities;

Temixco has 49 PCUs in one community; Xochitepec has 29

PCUs in three communities (Table 1). Official census data

from the Instituto Nacional de Estadistica, Geografia e

Informacion (INEGI, 1995) were used to stratify each

PCU by altitude and socio-economic category, defined by

high or low risk variables associated with housing quality.

The Cuernavaca metropolitan area rises from 1200 to

2200m altitude above mean sea level (m a.s.l.) on south-

ward-facing slopes of the Sierra Madre Occidental. Three

altitudinal categories were defined for our survey:< 1500,

1500–1800 and >1800m a.s.l. A socio-economic index was

developed using six variables recorded for 142 653 houses in

the most recent census (INEGI, 1995): wall materials, roof

materials, floor materials, piped water, kitchen inside

house, number of rooms used only for sleeping. Explan-

atory variables (wall, roof, floor materials, number of sleep-

ing rooms) were assigned numerical values 0, 1 or 2

representing three economic strata (0< 1< 2). Binary vari-

ables (presence of kitchen and piped water) were assigned

either 0 or 1 according to absence or presence. Responses

for all houses in each PCU were tabulated and the average

value used for socio-economic ranking of the PCU as rela-

tively good, standard, poor or very poor.

220 J. M. Ramsey et al.

# 2005 The Royal Entomological Society, Medical and Veterinary Entomology, 19, 219–228

Required sample sizes for houses of each stratum in each

PCU were calculated (assuming 5% infestation index deter-

mined by preliminary surveys) according to the proportions

of houses in each altitudinal and socio-economic category.

Using PCU maps, at least one housing block was selected

per PCU in a balanced spatial sample (square symbols on

Fig. 1). A maximum of three housing blocks was sampled

when PCU size was large. Selection procedure for the three

houses to be sampled involved on-site inspection of each

block, enumeration of the number of houses on each side of

the block, and selection of the middle house for each of

three sides. In a preliminary survey, householders were

given information regarding the study and about Chagas

disease and, if they agreed to involve their household, were

requested to sign informed consent. If a selected household

declined (only < 3% of initial selection), the house immedi-

ately to the right followed by the house on the left was

selected instead. Overall, 1026 houses in 342 blocks were

selected for the main survey.

Survey questionnaire

An adult family member from each house was asked

about demographic and socio-economic factors: personal

status, age, gender, residency, healthcare system, healthcare

giver, literacy, last year of school, occupation and employ-

ment record, sleep habits in farm areas, presence of coron-

ary or digestive disease, blood transfusion history; housing:

age of construction, number of days/week occupied, elec-

tricity, water source, number of rooms, sleeping arrange-

ments for family members, wild animals observed from

house, observation of bugs inside or outside house, domes-

tic use of insecticides, use of bednets, number of doorways,

windows that open, storeys (levels) of the house, materials

of roof, wall and floor, wall integrity, cleanliness, presence

of agricultural products (harvest) inside house, presence of

firewood inside house, presence and location of kitchen,

adjacent lots empty (unconstructed), type of surrounding

vegetation, size of lot and construction, presence of animals,

vermin (rodents), junk, stone piles, construction materials

or agricultural products in the yard, domestic animals

roaming freely, livestock in adjacent lots; and Chagas

disease-specific risk variables: capability to recognize

T. pallidipennis as a vector bug, having been bitten by

bug(s), collection of bugs, knowledge of Chagas disease,

and whether their blood had been tested for T. cruzi.

Spatial database for sample area and risk factors

Cartographic databases from CONABIO (http://

www.conabio.gob.mx) were downloaded into a universal

information system, including topography, city block map-

ping, roads, rivers and water causeways, administrative

divisions, demographic and socio-economic data for

PCUs. Georeferenced political and administrative data-

bases were obtained from the Instituto Nacional de

Estadistica, Geografia e Informacion (INEGI) and analysed

using MapInfo v.5 (MapInfo Corp., New York). Georefer-

ences were determined by GPS 320 receiver (Magellan

Corp., CA). The position (longitude, latitude) of each

house sampled was registered by housing code in the

system, and spatial variables associated with house position

such as altitude, distance from water causeways, distance

from farmland or periphery of the metropolitan area, and

infestation were analysed using buffer areas and bivariate

analysis.

Entomological evaluation

As for our previous survey of domestic Triatominae in

the State of Oaxaca (Ramsey et al., 2000), we employed

standard methods for sampling bugs and calculated the

infestation, colonization and crowding indices (WHO,

Table 1. Infestation indices for T. pallidipennis by county and community in the Cuernavaca metropolitan area. PCUs, number of population

catchment units.

Houses

Infestation index (%)

County Community PCUs sampled/total Community County

Cuernavaca Cuernavaca 110 427/75674 3.0 3.0

Jiutepec Jiutepec 61 252/28208 2.0

Calera Chica 2 6/931 16.7

Progreso 13 48/2732 4.2

Independencia 4 15/1387 13.3 3.1

Emiliano Zapata Emiliano Zapata 14 55/6039 1.8

Tezoyuca 1 6/763 16.7

Tres de Mayo 9 42/3210 7.1 4.9

Temixco Temixco 49 191/17819 6.8 6.8

Xochitepec Xochitepec 12 45/3226 4.4

Chiconcuac 2 12/1238 0.0

UHJMMP 15 30/1426 10.0 5.7

Triatoma pallidipennis risk in Mexico 221


1991). Results were simplified because T. pallidipennis was

the only species of bug found. During each house survey

visit by a pair of collectors, the niches indoors and outdoors

were searched for 20min each. We recorded the presence

and site of bugs (inside or outside the house), hour of

collection, and movement of furniture or outdoor materials

for bug searching. A flash-light was used to help see into

cracks and crevices throughout the fabric of buildings,

behind pictures on the walls, furniture, in closets and espec-

ially under bedding material. Note was made of bug faecal

streaks on walls or pictures (Schofield, 1994) and the

presence of bug exuviae. The intradomicile area was defined

as the space contained within the main walls of the dwelling

where household members slept. The peridomicile area was

defined as that area, outside the main walls of the dwelling,

which belonged to and/or was used by household members

for family activities; in most cases the peridomicile area was

surrounded by a fence. Surveying the peridomicile included

searching mounds of construction materials, agricultural

products, rocks, fence and areas with animals.

From each house, all the bugs collected alive and dead

(adults and nymphs), exuviae (cast nymphal skins), eggs

and eggshells were placed in plastic tubes (separately for

intra- and peridomicile samples) numbered with the house

Fig. 1. Socio-economic strata for PCUs (white¼ good, light grey¼ standard, dark grey¼poor, black¼poorest) from the Cuernavaca

metropolitan area (including the five counties of Cuernavaca, Jiutepec, Temixco, Emiliano Zapata, and Xochitepec), with position of sample

blocks (squares) and infested houses (dark circles) from population sample marked. Altitude ranges from > 2200m a.s.l. in the north, falling

to 1200ma.s.l. in southern Cuernava area (bottom of map): see http://www.maps-of-mexico.com/non-vr-cities/cuernavaca.html for further

details.



code, collection site and hour of capture. Bugs were trans-

ported daily to the laboratory, registered and identified

using the key of Lent & Wygodzinsky (1979). The numbers,

species and stage(s) of bug obtained from each house were

recorded on laboratory forms.

A second method for detection of infestation was the

collection of bugs opportunistically by the householders

themselves. After the timed manual search, the team gave

a tube or a self-sealing plastic bag to the householders,

requesting them to collect and keep any bugs found. The

tubes contained the house code, and the family was

requested to note the date and place of capture. House-

holders were asked to telephone the Chagas laboratory

should they find any bugs during the ensuing period of

6months. Samples of 10% of the houses were called after

6months to check if any bugs had been collected. After

identification, bugs collected by householders since the

pre-survey interview were included in house collection

records.

Bug infections

As described previously (Ramsey et al., 2000, 2003a),

samples of bug faeces were obtained (if necessary by press-

ing on the bug abdomen to force defaecation), diluted in

saline solution and examined microscopically at 400� for

presence of T. cruzi flagellates.

Data analysis

An infested house was defined as any house where at least

one bug or exuvium was found by any sampling method.

From results of timed searches, the standard entomological

indicators were calculated (WHO, 1991) to express the

infestation index (houses infested/examined), density index

(total bugs/houses), crowding index (bugs/infested house)

and colonization index (percentage of infested houses with

bug nymphs) for each county, community, altitude and

socio-economic stratum. All potential risk factors were sub-

mitted to univariate and bivariate analysis (Mantel–Haentzel)

and significant variables were submitted to multivariate

logistic regression for association with infestation using

STATA, version 6 (http://www.stata.com). Possible confusing

variables were assessed with dummy models, which were

reduced to include only those variables with significant

associations.

Results

Entomological indices

All the triatomine bugs obtained from searches in and

around 1129 houses were identified as T. pallidipennis. The

proportion of infested housing was 4.1% overall in the

metropolitan area, ranging from only 3% in central

Cuernavaca up to 6.8% in the southern suburb of Temixco

and nearly 17% in some communities of Juitepec and

Emiliano Zapata counties (Table 1). The infestation index

was negatively correlated with both altitude (Table 2, Fig. 2)

and socio-economic stratum (Table 3, Fig. 3): there were

very significantly more infestations in poorer PCUs, both

indoors and peridomiciliary (P< 0.001), and these housing

conditions were associated with lower altitudes (Fig. 1).

Conversely the colonization index (percentage of infested

houses having immature bugs – indicating a breeding infes-

tation) correlated positively with altitude (Table 2, Fig. 2),

but apparently not with socio-economic category of the

PCU (Table 3). Bug nymphs were found in 21% of infested

houses overall, ranging from 19.2% in houses situated

below 1500m altitude to 50% of the infested houses above

1800m altitude (Table 2). Even so, the crowding index,

representing bug density per infested house, correlated

inversely with socio-economic category (Table 3) and with

altitude (Fig. 4), especially for nymphal stages (P< 0.0001)

representing 85% of bugs found in the lowest and poorest

PCUs, compared with 20% and 50% nymphs among infes-

tations in good and standard PCUs, respectively. Hence the

bug crowding index was> 4-fold greater in very poor/poor

PCUs compared with good/standard strata.

The density index (a measure of the clustering/dispersion

of infested houses) rose towards the middle of the metro-

politan area (spanning 1500–1800m altitude, Table 2) due to

the greatest density of housing, and was strongly correlated

with poorer socio-economic strata (Table 3).

Infection rates

Trypanosoma cruzi infection was detected in 47.9% and

51.9% of T. pallidipennis collected from intradomicile and

peridomicile situations, respectively, in the metropolitan

area. Infection rates were significantly lower in nymphs

than adults; the highest rate was in male bugs outdoors

(Table 4).

Risk factors associated with infestation

Bivariate analysis showed that several variables of the

peridomicile and surrounding-area were significantly asso-

ciated with infestation. The leading environmental factors

were: houses constructed recently (0–7 years), showing

2.8-fold (95% CI¼ 1.4–5.7) greater risk of infestation, as

Table 2. Entomological indices for T. pallidipennis in the Cuerna-

vaca metropolitan area by altitudinal strata.

Altitude (m)

Infestation

index (%)

Colonization

index (%)

Crowding

index

Density

index (%)

< 1500 5.4 19.2 6.5 13.2

1500–1800 3.1 35.7 2.3 22.7

> 1800 2.2 50.0 4.0 15.4



compared with older houses; and presence of at least one

empty adjacent lot, representing 4.3-fold increase in risk of

infestation (95% CI¼ 2.3–8.0). Larger garden size

(¼ 80m2), peridomiciliary presence of livestock (pigs >sheep > cows > horses), rodents, firewood or harvest pro-

ducts, presence of grassland or forest around the house were

all significantly associated with greater risk of infestation.

Householder reports of triatomines correlated well

(85.4% agreement, K¼ 0.2547) with results of our timed

searches for infestation (OR¼ 21.5, 95% CI¼ 10.5–44.2).

Therefore, to augment the findings (and confidence inter-

vals), we combined the data for all households where bugs

were collected (by us) or reported (by residents) inside or

outside the house. Thus, in addition to risk factors identi-

fied in the preceding paragraph, the presence of cats and

dogs, plus factors such as altitude and PCU category were

found to be associated with infestation (Table 5). Structural

features with significant odds ratios included houses with

fewer doors and built with only one level (bungalow), roof

materials, less cleaning and presence of pet animals.

In multivariate analysis, none of the housing construction

variables remained significant (Table 6). Among the perido-

miciliary factors, garden size (OR¼ 1.7, 95% CI¼ 1.2–2.4)

and presence of sylvan animals, specifically opossums

(OR¼ 2.02, 95% CI¼ 1.4–3.0) and squirrels (OR¼ 2.1,

95% CI¼ 1.2–3.6), remained significant. Dogs ranging

freely in and out of the house was also significant. Empty

adjacent lots and the presence of pigs roaming in surround-

ing areas were of great significance. Lower altitude

remained very significantly associated with infestation,

while the PCU socio-economic category was lost as a con-

founder.

Spatial analysis

Environmental factors analysed for correlation with

infestation were distance to river and water causeways (buf-

fer zones of 100, 200, 400, 800,> 800m) and distance from

the peripheral border of the metropolitan area (200, 400,

600, 800m). Neither of these two variables was significantly

associated with infestation.

Knowledge of triatomines and Chagas disease

Proportions of householders who claimed to ‘know’ tri-

atomine bugs were 43% from non-infested and 89% from

infested houses (Table 7), representing 45% of the total

population. Of these, only 29% and 78%, respectively,

were able to correctly identify T. pallidipennis nymphs or

adults. Hence, only 33% of the population claiming to

‘know’ bugs, representing 16% of the total population,

0

1

2

3

4

5

6

<1500 1500–1800 >1800

Altitude

Infe

stat

ion

inde

x (%

)

0

10

20

30

40

50

60

Colonization index (%

)

Infestation Colonization

Fig. 2. Infestation and colonization indices in the Cuernavaca

metropolitan area according to altitude strata.

Table 3. Entomological indices for T. pallidipennis in the Cuerna-

vaca metropolitan area by socio-economic strata.

Socio-economic

stratum

Infestation

index (%)

Colonization

index (%)

Crowding

index

Density

index (%)

Good 2.0 25.0 3.3 6.4

Standard 3.6 33.3 1.7 14.1

Poor 5.4 21.4 6.6 20.5

Very poor 6.5 20.0 8.1 25.6

1.5

2.5

3.4

4.2

0.5

1.2

2

3

00.5

11.5

22.5

33.5

44.5

Good Standard Poor Poorest

PCU

Infe

stat

ion

inde

x (%

)

Intradomicile Peridomicile

Fig. 3. Intradomicile and peridomicile infestation indices by socio-

economic strata in the Cuernavaca metropolitan area.

0

5

10

15

20

25

30

Cro

wdi

ng in

dex

<1500 1500–1800 >1800

Altitude (masl)

Nymphs Males Females

Fig. 4. Triatoma pallidipennis crowding indices in houses of the

Cuernavaca metropolitan area by altitudinal strata.



really could identify bugs correctly. Householders who

reported receiving, or having heard of information regard-

ing Chagas disease occupied 15% of non-infested but 33%of infested houses, comprising 15% of the total population.

Discussion

Despite the long history of triatomine presence in the Cuer-

navaca metropolitan area (Brumpt et al., 1939), this is the

first effort to quantify infestations, revealing the over-

whelming predominance of T. pallidipennis and identifying

risk factors. By microscopic examination of bug faeces,

T. cruzi infections were found in about half of the T. pallidi-

pennis sampled from both intra- and peridomiciliary bio-

topes, but no salivary infections of Trypanosoma rangeli

were detected. Bautista et al. (1999) reported generally

lower rates of T. cruzi infection in T. pallidipennis across

Morelos (domestic 29%, peridomestic 4%, sylvan 20%),

whereas Vidal-Acosta et al. (2000) reported 50%,

Herlindo-Jaimes (1998) reported 54%, Arriola (1989)

reported 81% and Cortes-Jimenez et al. (1996) reported

88% infection rate of T. pallidipennis with T. cruzi in

Cuernavaca, together with 70% T. cruzi in some T. barberi

collected there. Although Cuernavaca is the type-locality of

T. barberi, this bug species is more abundant in other parts

of Morelos and other States of Mexico (Zarate & Zarate,

1985; Ramsey et al., 2000, 2003a), usually associated with

Neotoma woodrats (Peterson et al., 2002). Both T. barberi

and T. pallidipennis are restricted to Mexico and regarded as

locally important vectors of Chagas disease (Lent &

Wygodzinsky, 1979; Tay, 1980;Martınez Licona&Contreras

Zavala, 2003; Ramsey et al., 2003b, 2005; Enger et al., 2004).

Houses infested with T. pallidipennis were found across

the entire metropolitan area of Cuernavaca, where at least

150 000 structures are inhabited by perhaps 3/4 million

people, most PCUs having 2% to 7% houses infested,

some with almost 17% infestation index. Elsewhere in wes-

tern Mexico, T. pallidipennis predominates in many com-

munities (Espinoza-Gomez et al., 2002; Ramsey et al.,

2003a; Enger et al., 2004), usually not excluding other tri-

atomines as in Cuernavaca. Although the colonization index

was proportional to altitude, perhaps due to bugs avoiding

lower ambient temperatures outside, the only thriving intra-

domiciliary infestations (with high nymphal crowding

indices) occurred below 1500 ma.s.l. in socio-economically

poorer PCUs of four counties (Emiliano Zapata, Temixco,

Xochitepec and most of Jiutepec) where risks were greatest.

The median risk area is composed of PCUs between 1500

and 1800 ma.s.l. in Cuernavaca county, plus a small part

(�5%) of Jiutepec, where infestation may be augmented by

passive transport of bugs from lower areas. The lowest risk

area in metropolitan Cuernavaca is that part of Cuernavaca

county located above 1800 ma.s.l. Thus, considering the

generally positive association between altitude and socio-

economically better houses in Cuernavaca (contrary to the

hillside favela slums of cities such as Rio de Janeiro), risks

of T. pallidipennis infestation declined from 5.4% in PCUs

below 1500ma.s.l., to 3.1% at intermediate altitudes, to

only 2.2% in suburbs above 1800ma.s.l. Hence, the popula-

tion at risk in Cuernavaca metropolitan areas was estimated

to be 612 000 people.

More than 70% of the triatomines collected in this study

were found by householders, not by timed entomological

searches. The sensitivity of searching for bugs as a method

for detection of infestation was minimal in this urban situa-

tion where crowding indices are so low, although it may be

an important complement to householder collections. As a

way of teaching community members to observe where bugs

often hide, both inside and outside houses, searches have

some dividends.

Within each community defined by environmental fac-

tors, the risk of exposure to vectors of Chagas disease is not

based simply on the socio-economic category of the PCU.

For example, a house in a low socio-economic category

PCU (potential high risk) within a low risk area, probably

has no greater risk of infestation than for a higher PCU

category within the same risk area. This is because major

determinants of risk are derived from the environs (e.g.

altitude, adjacent lots), irrespective of the house character-

istics. From our surveys, the most important factor asso-

ciated with T. pallidipennis infestation is the presence of

unconstructed (empty) lots of land adjacent to the house,

and the risk doubles with each additional empty lot. This

exposure undoubtedly comes from animals in such habitats.

Accordingly, the reported presence of domestic animals

(particularly roaming dogs), livestock (especially foraging

pigs) or arborial and burrowing animals such as squirrels

and opossums (Didelphis), was significantly associated with

infestation. Elsewhere in Morelos, T. pallidipennis is known

to inhabit rock piles and rodent nests (Baiomys, Liomys,

Neotoma, Peromyscus) and caves with bats (Artibeus,

Balantiopteryx, Glossophaga, Pteronotus, Sturnira) that

carry T. cruzi (Bautista et al., 1999; Villegas-Garcıa &

Santillan-Alarcon, 2001; Ramsey et al., 2003a). Moreover,

street-collected dogs in Cuernavaca city have 2.3% seropre-

valence for T. cruzi (J. Ramsey, unpublished data), presum-

ably transmitted by T. pallidipennis. The bugs probably rely

on these hosts, as well as vermin such as Mus musculus and

Rattus norvegicus, to maintain urban populations of triato-

mines, presumably bringing transmission risks of zoonotic

T. cruzi. The association of the size of the peridomicile

garden with infestation again points to the peridomestic

Table 4. Trypanosoma cruzi. infection rates of T. pallidipennis

from houses in Cuernavaca.

Biotope (%)

Stage/sex of bug Intradomicile Peridomicile

Nymph 37.0 33.3

Female 54.2 44.2

Male 55.0 66.7

Total 47.9 51.9



maintenance of triatomine populations: the larger the gar-

den, the greater the probability of sylvan animal nests

undetected by inhabitants. We had expected that proximity

to major thoroughfares of sylvan animals, such as the water

causeways, would confer more exposure to the surrounding

houses. Since we did not find a spatial association between

infestation and this environmental factor, nor the periphery

of the urban area, this points to the importance of peri-

domestic animals (livestock, pets, vermin and sylvan

intruders such as opossums and squirrels) rather than

those which burrow in truly sylvan habitats.

Among intradomicile factors related to risk of exposure,

due to bug infestation, the most significant was the unrest-

ricted movement of dogs between indoors and outdoors.

Some construction variables (materials for walls, roof,

floor, water source, electricity, construction area, numbers

of doors or windows, levels of the house, number, and type

of sleeping furniture) and factors due to the inhabitants

(numbers and duration of residents, use of mosquito netting

and/or insecticides) seemed to be significant with bivariate

analysis, but none of them remained significant with

multivariate analysis. This is not surprising, given the high

Table 5. Odds Ratio for risk factors significantly associated with infestation in bivariate analysis

Houses

Factor

Infested

No. (%)

Non-infested

No. (%)

Odds Ratio

(95% CI)

Environmental

Altitude (m) > 1800 9 (10.1) 80 (89.9) 1

1500–1800 55 (10.2) 484 (89.8) 1.01 (0.48–2.12)

< 1500 137 (27.3) 364 (72.7) 3.35 (1.63–6.85)

PCU category Good 27 (9.3) 262 (90.7) 1

Standard 64 (15.8) 342 (84.2) 1.82 (1.12–2.93)

Poor 65 (22.3) 227 (77.7) 2.78 (1.71–4.5)

Very poor 45 (31.7) 97 (68.3) 4.50 (2.65–7.66)

Internal house structure

Age since construction > 15 year 8 (2.5) 314 (97.5) 1

8–15 year 12 (3.2) 358 (96.8) 1.31 (0.5–3.25)

1–7 year 24 (6.8) 330 (93.2) 2.80 (1.4–5.7)

House storeys 2 levels 41 (13.2) 270 (86.8) 1

1 level 159 (19.5) 657 (80.5) 1.59 (1.1–2.31)

Number of doors > 4 78 (14.1) 476 (85.9) 1

1–3 117 (21.5) 427 (78.5) 1.67 (1.22–2.29)

Dog sleeps indoors 21 (28.0) 54 (72.0) 1.86 (1.1–3.16)

Cat sleeps indoors 16 (29.1) 39 (70.9) 1.95 (1.07–3.56)

Peridomicile

Garden size ¼ 80m2 124 (23.0) 416 (77.0) 1.98 (1.45–2.71)

Dog resident 129 (22.6) 442 (77.4) 1.97 (1.43–2.7)

Cat resident 66 (25.0) 198 (75.0) 1.80 (1.29–2.51)

Dog visits freely 106 (25.8) 305 (74.2) 2.28 (1.67–3.13)

Cat visits freely 43 (25.7) 124 (74.3) 1.74 (1.18–2.57)

Opossum(s) seen 71 (30.6) 161 (69.4) 2.60 (1.86–3.63)

Rodent(s) seen 82 (21.4) 301 (78.6) 1.43 (1.05–1.96)

Squirrel(s) seen 33 (35.1) 61 (64.9) 2.78 (1.77–4.38)

Rock mound(s) 81 (24.2) 254 (75.8) 1.76 (1.28–2.42)

Firewood 13 (8.8) 145 (91.2) 2.51 (1.29–4.89)

Harvest 8 (9.0) 81 (91.0) 2.57 (1.16–5.69)

Surrounding area

> 1 empty lot 109 (28.5) 274 (71.5) 2.83 (2.07–3.86)

1 empty adjacent lot 64 (24.0) 203 (76.0) 2.24 (1.57–3.19)

2 empty adjacent lots 30 (34.5) 57 (65.5) 3.74 (2.29–6.13)

3 empty adjacent lots 14 (53.8) 12 (46.2) 8.29 (3.72–18.48)

Crops 3 (20.0) 12 (80.0) 1.52 (0.42–5.47)

Grass 83 (23.2) 274 (76.8) 1.84 (1.33–2.54)

Forest 14 (31.1) 31 (68.9) 2.74 (1.41–5.33)

Horse(s) 49 (27.5) 129 (72.5) 1.99 (1.37–2.88)

Cow(s) 30 (31.9) 64 (68.1) 2.37 (1.49–3.77)

Sheep 18 (36.0) 32 (64.0) 2.76 (1.51–5.02)

Pig(s) 11 (37.9) 18 (62.1) 2.90 (1.35–6.24)



variability of construction types in an urban area; such

factors are associated with infestations in rural communities

(Starr et al., 1991; de Andrade et al., 1995; Ramsey et al.,

2000; Enger et al., 2004).

Overall, 45% of households reported ‘knowing’ triato-

mines, the proportion for occupants of infested houses

being double that of non-infested houses. However, only

33% of these informants (representing 15% of households)

correctly identified specimens of Triatoma shown to them.

Moreover, only 16% of all households had heard of, or had

received information regarding Chagas disease: again this

score was double for infested households. These data reflect

unacceptably low knowledge of this insidious disease and its

vector in the metropolitan area. Even so, residents were

reliable informants for reporting the presence of bugs inside

or outside houses, with 85% correlation between reported

bug sightings and detection of infestation. Therefore, com-

munity health education on this topic is predicted to have a

high impact for improving triatomine surveillance and con-

trol, consequently reducing T. cruzi transmission risks and

preventing Chagas disease incidence at Cuernavaca, with

potential to serve as a model municipality programme for

community participation in the campaign to eliminate this

problem at the State (SSM, 2001), national (SSA, 2001) and

regional levels (WHO, 1991; Dias et al., 2002; Ramsey &

Schofield, 2003).

Acknowledgements

We are grateful to the Mayors of Cuernavaca, Jiutepec,

Temixco, Emiliano Zapata and Xochitepec for their

support and assistance throughout this study. Financial

support was given by the Mexican Ministry of Health

(INSP #100191) and the European Community Latin

American Triatomine Network (ECLAT).

References

de Andrade, A.L., Zicker, F., De Oliveira, R.M., Da Silva,

I.G., Silva, S.A., de Andrade, S.S. & Martelli, C.M. (1995)

Evaluation of risk factors for house infestation by Triatoma

infestans in Brazil. American Journal of Tropical Medicine and

Hygiene, 53, 443–447.

Arriola, G.M.T. (1989) Frecuencia de Trypanosoma cruzi (Kine-

toplastida: Trypanosomidae) en Triatoma pallidipennis (Hemi-

ptera: Reduviidae) recolectadas en la ciudad de Cuernavaca,

Morelos. Resumenes i Encuentro Estatal sobre Entomologıa

Medica y Veterinaria y il Jornadas Medicas de Salud en el

Trabajo, Cuernavaca, Morelos, Mexico, pp. 109–115.

Bautista, N.L., Garcia de la Torre, G.S., de Haro Arteaga, I. &

Schettino, P.M.S. (1999) Importance of Triatoma pallidipennis

(Hemiptera: Reduciidae) as a vector of Trypanosoma cruzi

(Kinetoplastida: Trypanosomatidae) in the State of Morelos,

Mexico, and possible ecotopes. Journal of Medical Entomology,

36, 233–235.

Brumpt, E., Mazzoti, L. & Brumpt, L.C. (1939) Enquete

epidemiologiques sur la maladie de C. Chagas au Mexique.

Reduvides vecteurs, animaux reservoirs de virus, cas humains.

Annales de Parasitologie, 17, 299–312.

Cortes-Jimenez, M., Nogueda-Torres, B., Alejandro-Aguilar, R.,

Isita-Tornell, I. & Ramırez-Moreno, E. (1996) Frequency of

triatomines infected with Trypanosoma cruzi collected in

Cuernavaca city, Morelos, Mexico. Revista Latinoamericana de

Microbiologia, 38, 115–119.

Dias, J.C.P., Silveira, A.C. & Schofield, C.J. (2002) The impact of

Chagas disease control in Latin America. Memorias do Instituto

Oswaldo Cruz, 97, 603–612.

Enger, K., Ordonez, R., Wilson, M.L. & Ramsey, J.M. (2004)

Evaluation of risk factors for rural infestation by Triatoma

pallidipennis (Hemiptera: Triatominae), a Mexican vector of

Chagas disease. Journal of Medical Entomology, 41, 760–767.

Espinoza-Gomez, F., Maldonado-Rodrıguez, A., Coll-Cardenas,

R., Hernandez-Suarez, C.M. & Fernandez-Salas, I. (2002)

Presence of Triatominae (Hemiptera, Reduviidae) and risk of

transmission of Chagas disease in Colima, Mexico. Memorias do

Instituto Oswaldo Cruz, 97, 25–30.

Guzman, C., Garcıa, L., Verdugo, J., Guerrero, S., Torres, M.,

Ramirez, C. & Velasco, O. (1998) Riesgo de transmision de

Trypanosoma cruzi por transfusion de sangre en Mexico. Revista

Panamericana de Salud Publica, 4, 94.

Herlindo-Jaimes, F. (1998) Infeccion de triatominos y presencia de

Trypanosoma cruzi en mamıferos silvestres y peridomesticos del

municipio de Jiutepec, Morelos, Mexico. Tesis de Licenciatura,

Facultad de Ciencias Biologicas, Universidad Autonoma del

Estado de Morelos, Cuernavaca, Morelos, Mexico.

INEGI (1995) Annuario Estadistico del Estado de Morelos. Instituto

Nacional de Estadistica, Geografia e Informacion, Mexico.

Table 6. Odds Ratio (95% confidence interval, CI) of risk factors

associated with infestation in multivariate regression analysis

OR (95% CI)

Factor Crude Adjusted

Environmental

Altitude < 1500m 3.35 (1.63–6.85) 2.98 (2.07–4.29)

Peridomicile

Garden size > 80m2 1.98 (1.45–2.71) 1.68 (1.18–2.38)

Dog entering house 2.28 (1.67–3.13) 1.50 (1.06–2.12)

Opossum near house 2.60 (1.86–3.63) 2.02 (1.38–2.95)

Squirrel near house 2.78 (1.77–4.38) 2.12 (1.24–3.61)

Surrounding area

With 1 empty lot 2.24 (1.57–3.19) 2.09 (1.42–3.06)

With 2 empty lots 3.74 (2.29–6.13) 2.33 (1.35–4.02)

With 3 empty lots 8.29 (3.72–18.48) 6.25 (2.55–15.3)

Presence of pigs 2.90 (1.35–6.24) 3.53 (1.49–8.34)

Table 7. Proportion of the infested and non-infested households

that had received information regarding Chagas disease or

triatomines

Questionnaire topic Not infested Infested

Number of houses 984 (95.9%) 42 (4.1%)

Report ‘knowing’ bugs (%) 43.1 89.1

Proportion correctly identifying bugs (%) 28.8 78.0

Adjusted ‘know’ bugs (%) 12.4 69.5

Have received information

regarding Chagas disease (%)

15.1 32.6



Lent, H. & Wygodzinsky, P. (1979) Revision of the Triatominae

(Hemiptera, Reduviidae) and their significance as vectors of

Chagas disease. Bulletin of the American Museum of Natural

History, 163, 127–520.

Martınez Licona, F.S. & Contreras Zavala, C. (2003) La

Enfermedad de Chagas en Mexico: Situacion de Rhodnius

prolixus. Reunion Internacional para el Establecimiento de

Criterios de Certificacion de la Eliminacion de Rhodnius prolixus

(Guatemala City, 5–7 March 2003), pp. 12–16. Document OPS/

DPC/CD/245/03. Pan–American Health Organization – Orga-

nizacion Panamericana de la Salud (PAHO–OPS), Washington,

D.C. (In Spanish).

Moreno Lopez, R.M., Sanchez Paredes, L., Munoz Jimenez, L.,

Monteon, V.M. & Rees Lopez, P.A. (2001) Cardiopatıa

chagasica en Tehuantepec. Informe preliminar. Archivos de

Cardiologia de Mexico, 71, 43–49.

Peterson, A.T., Sanchez-Cordero, V., Beard, C.B. & Ramsey, J.M.

(2002) Ecologic niche modeling and potential reservoirs for

Chagas disease, Mexico. Emerging Infectious Diseases, 8, 662–667.

Ramsey, J.M., Cruz-Celis, A., Salgado, L., Espinosa, L., Ordonez,

R., Lopez, R. & Schofield, C.J. (2003a) Efficacy of pyrethroid

insecticides against domestic and peridomestic populations of

Triatoma pallidipennis and Triatoma barberi (Reduviidae:

Triatominae) vectors of Chagas’ disease in Mexico. Journal of

Medical Entomology, 40, 912–920.

Ramsey, J.M., Ordonez, R., Cruz-Celis, A., et al., (2000)

Distribution of domestic Triatominae and stratification of

Chagas disease transmission in Oaxaca, Mexico. Medical and

Veterinary Entomology, 14, 19–30.

Ramsey, J.M., Ordonez, R., Tello Lopez, A., Pohls, J.L., Sanchez-

Cordero, V. & Peterson, A.T. (2003b) Chagas disease in Mexico:

morbidity, mortality, risk areas and disease burden. (Abstract)

XXX Annual Meeting on Basic Research in Chagas Disease –

XIX Meeting of the Brazilian Society of Protozoology,

Caxambu, MG, Brasil, 10–12 November 2003. Revista do

Instituto de Medicina Tropical de Sao Paulo, 45 (Suppl. 13), 198.

Ramsey, J.M., Ordonez, R., Tello-Lopez, A., Pohls, J.L., Sanchez-

Cordero, V. & Peterson, A.T. (2005) Chagas Disease in Mexico:

morbidity, mortality, transmission risk areas and disease

burden. Emerging infections Diseases, 11 in press.

Ramsey, J.M. & Schofield, C.J. (2003) Control of Chagas disease

vectors. Salud Publica de Mexico, 45, 123–128.

Schmunis, G. (1996) La Tripanosomiasis americana como

problema de salud publica. Pan American Health Organization

Publication No, 547, 3–31.

Schofield, C.J. (1994) Triatominae Biology and Control. Euro-

communica Publications, Bognor Regis, U.K.

SSA (2001) Secretarıa de Salud de Mexico. Norma Oficial

Mexicana Para la Vigilancia, Prevencion y Control de

Enfermedades Transmitidas por Vector (PROY-NOM-032-

SSA2-2000). Diario Oficial de la Federacion, 565, 5, Section 2,

1–45.

SSM (2001) 2001 Programa de Actividades para la Prevencion de la

Enfermedad de Chagas en Morelos. Secretaria de Salud del

Estado de Morelos, Cuernavaca, Mexico.

Starr, M.D., Rojas, J., Zeledon, R., Hird, D. & Carpenter, T.

(1991) Chagas’ disease: Risk factors for house infestation by

Triatoma dimidiata, the major vector of Trypanosoma cruzi in

Costa Rica. American Journal of Epidemiology, 133, 740–747.

Tay, J. (1980) La enfermedad de Chagas en la Republica Mexicana.

Revista de Salud Publica Mexicana, 22, 409–450.

Vidal-Acosta, V., Ibanez-Bernal, S. & Martınez-Campos, C. (2000)

Infeccion natural de chinches Triatominae con Trypanosoma

cruzi asociadas a la vivienda humana en Mexico. Salud Publica

de Mexico, 42, 496–503.

Villegas-Garcıa, J.C. & Santillan-Alarcon, S. (2001) Sylvatic focus

of American Trypanosomiasis in the State of Morelos, Mexico.

Revista de Biologıa Tropical, 49, 685–688.

WHO (1991) Control of Chagas Disease. Technical Report Series

811. World Health Organization, Geneva.

Zarate, L.G. & Zarate. R.J. (1985) A checklist of the Triatominae

(Hemiptera: Reduviidae) of Mexico. International Journal of

Entomology, 27, 102–127.

Accepted 18 January 2005



Risk factors associated with house infestation by the Chagas disease vector Triatoma pallidipennis...

Documents

Transcript of Risk factors associated with house infestation by the Chagas disease vector Triatoma pallidipennis...