Canine Leishmaniasis in Southeastern Spain

To examine prevalence changes and risk factors for canine leishmaniasis, we conducted a cross-sectional seroprevalence study and a survey during April–June 2006. Seroprevalence had increased at the meso-Mediterranean bioclimatic level over 22 years. Risk was highest for dogs that were older, large, lived outside, and lived at the meso-Mediterranean level.

I t has been suggested that climate change has the potential to change the transmission intensity of vector-borne diseases such as leishmaniasis, but supporting literature is lacking (1,2). Because long-term quality data on leishmaniasis caused by Leishmania infantum and its vector (3)(4)(5)(6)(7)(8)(9) are available for the Alpujarras region of southeastern Spain (Figure 1), this is an ideal area for studying changes in the prevalence of canine leishmaniasis in a changing environment. Our study objectives were to determine whether any changes had occurred in the prevalence of canine leishmaniasis over 22 years and to identify risk factors for this disease.

The Study
To achieve the fi rst objective, we conducted a crosssectional study in the Alpujarras from April through June 2006. We then compared current leishmaniasis seroprevalence data with data from 2 cross-sectional surveys conducted in 1984 and 1991 (3)(4)(5). The villages sampled for all 3 studies were similar and had been selected at random from within each of the 3 bioclimatic levels (thermo-, meso-, and supra-Mediterranean) that comprise the inhabited zone of the Alpujarras (Figure 1) (10). For each level, respectively, altitudes are 0-700, 600-900, and 900-1,800 m above sea level; annual mean temperatures are 17-19, 13-17, and 8-15°C; and annual rainfalls are 200-350, 600-1,000, and 1,000-1,600 inches. The dates for sample collection were set to coincide with organized events at which dogs were gathered (e.g., antirabies vaccination campaigns). All 3 surveys used indirect immunofl uorescence for diagnosis; protocols and positivity threshold were identical. Dogs with a titer >160 were considered positive. To determine and compare the existence or lack of statistically signifi cant differences between present and past prevalence rates, we used the χ 2 or Fisher exact test.
To achieve the second objective, we conducted a survey. The owners of the dogs included in the cross-sectional study conducted in the Alpujarras from April to June 2006 were asked to complete an epidemiologic record for each dog tested; data on the animal and its environment were recorded for subsequent use in univariate and multivariate logistic regression analyses (Tables 1, 2). Density data were included for 2 vectors, Phlebotomus perniciosus and P. ariasi sandfl ies, captured with sticky traps in June 2006 in the same villages in which the surveys were conducted.
Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 15  No retrospective entomologic analysis was performed as had been done for canine leishmaniasis. In our 2006 survey, 57 (13.0%) of 439 dogs had an antibody titer >160 (seroprevalence rate), 268 (61.0%) dogs had titers 20-80, and 114 (26.0%) had no titer. In terms of bioclimatic level, canine leishmaniasis seroprevalence was 20.1% at the meso-Mediterranean, 13.3% at the thermo-Mediterranean, and only 1.1% at the supra-Mediterranean levels. Statistically signifi cant differences (χ 2 test p<0.001) indicate that these differences are not random. The evolution-in-time analysis ( Figure 2) shows how over 22 years  progressively increased at the meso-Mediterranean level, climbing from 9.2% in 1984 to 15.4% in 1991 and fi nally to 20.1% in 2006 (p = 0.015); in contrast, no signifi cant changes have taken place in global prevalence or in the other 2 bioclimatic levels studied (3-5; this study). Dogs at greatest risk for canine leishmaniasis in the disease-endemic region of the Alpujarras were large dogs (>25 kg) and older dogs (>4 years) that worked as guard dogs or sheepdogs, slept outdoors, and lived at the thermo-or meso-Mediterranean level in a village such as Torvizcón, where the P. perniciosus density is >4 sandfl ies/m 2 . Risk for these dogs was 54,571× greater than for dogs that were kept as pets, were small, were <4 years of age, and slept inside a house in a village such as Mecina Bombarón or Pórtugos at the supra-Mediterranean level where P. perniciosus density is <4 sandfl ies/m 2 .

Conclusions
Among the drivers of global change that have the potential to infl uence vector-borne diseases, climatic and nonclimatic (socioeconomic, demographic, and environmental) factors have been cited (1,2). Although it is not easy to attribute our fi ndings-progressive increase in seroprevalence at the meso-Mediterranean level and drop and subsequent rise at the thermo-Mediterranean level (Figure 2)-to 1 or more drivers, we can attempt to fi nd some form of association in the changes that have occurred in the Alpujarras throughout these 2 decades. Studies on climate change in Spain confi rm a warming tendency (refl ected at the global level), which provides evidence that temperatures have been increasing for a quarter of a century (11,12). Thus, we must assume that changes in temperature, rainfall, or humidity will have equally affected the 3 bioclimatic levels researched in the Alpujarras and may have infl uenced the spatial and temporal distribution and the seasonal dynamics of sandfl ies. An increase was detected in the period of P. perniciosus activity in the region; these effects were probably more notable at the meso-Mediterranean level because this is where the density of this vector species is at its highest (6; Martín-Sánchez et al., unpub. data). During the time period researched, the human population remained  constant in the Alpujarras; it has, however, progressively decreased in the disperse populations and increased in population centers. Unfortunately, no offi cial fi gures for the canine population are available, although we were informed by staff of local town halls that in 1991, the 615 dogs analyzed accounted for ≈100% of the total number of dogs registered in the villages sampled (3). The Alpujarras economy is based mainly on agriculture and livestock. One change during the period studied was the gradual disappearance of livestock enclosures in rural population centers. These enclosures appear to create a substantial risk for infection with Leishmania spp. (13,14) and are ideal places for sandfl y blood sucking, mating, and oviposition (15). With respect to risk factors, the increase in canine leishmaniasis seroprevalence as animal's age increases seems logical because in a leishmaniasis-endemic area, the greater the age, the longer the animal will have been exposed to sandfl ies and the greater the probability of having been bitten by an infected female sandfl y. Seroprevalence of canine leishmaniasis also increases gradually with weight, which could be attributed to the vector being more attracted to larger animals. Dogs that sleep outdoors are at greater risk than those sleeping indoors because the density of the vector (P. perniciosus) is greater outdoors than inside a house. The association between P. perniciosus density and canine leishmaniasis seems logical, considering that P. perniciosus is the main vector species in the Alpujarras. This is the only species found to have been infected in this region (8,9), although P. ariasi is a proven vector in the nearby region of the Axarquia (13).
In the Alpujarras, the percentage of dogs kept as pets has increased from 42% in 1991 (3) to 59% in 2006 (p<0.001). This increase corresponds to a general increase in standard of living and, along with other factors such as the disappearance of livestock enclosures from within population centers, would have acted to reduce transmission. Only the extension of the vector's activity period, which was detected at some trap sites, would have acted to increase transmission. Such extension of activity may be related to the increase in temperature brought about by climate change (11,12).
The Ministerio de Educación y Ciencia, Madrid (AGL2004-06909-C02-02), and the Junta de Andalucía (group CVI176) provided fi nancial assistance, and Emerging Diseases in a changing European eNvironment (EDEN; contract no. 010284, Integrated Project GOCE) awarded a PhD student grant to S.B. This publication is catalogued by the EDEN Steering Committee as EDEN0137 (www.eden-fp6project.net).
Dr Martín-Sánchez is a professor in the Parasitology Department of the University of Granada. Her main area of research is the epidemiology of leishmaniasis.