Landscape epidemiology of mosquito-borne canine heartworm (Dirofilaria immitis) in northern California, USA. I. Community-based surveys of domestic dogs in three landscapes.

Abstract

Endemic patterns of canine filariasis in northern California were studied in 3 communities differing in ecological landscape: Tehama City (riparian-agricultural), Palo Cedro (oakpine woodland), and Cedar Ridge (conifer-oak forest). Dog populations were surveyed by household for canine-infecting filariids; both microfilaremic and amicrofilaremic patent Dirofilaria immitis infections were documented (Knott, IFA/ELISA tests). Of 403 dogs resident in these communities, 111 were on preventive diethylcarbamazine (DEC) citrate therapy (55% tested), and 292 had no recent DEC history (87% tested). Dirofilaria immitis was highly endemic in all communities, with point prevalence rates of 34.3% (Tehama City), 28.0% (Cedar Ridge), and 21.1% (Palo Cedro). Amicrofilaremic occult infections constituted 14–29% of total cases. The frequency distribution of microfilaremic densities in community dogs was highly skewed (median = 4040 mf/ml). Dipetalonema reconditum microfilariae were found in 4.9% of dogs in the Tehama community. Significant heartworm risk factors were Residence period, exposure, and SEX as determined from stepwise logistic regression analysis. Heartworm risk increased with length of residence, compared to a reference population resident ≤ 1 year. Adjusted odds ratios were 3.1 in dogs resident 1–3 years, 5.5 (3–5 years), and 12.2 (>5 years) (98.7% confidence). Dogs kept strictly outside were at 5.4× greater risk of patent heartworm disease than animals kept indoors for at least part of the 24-h day (97.5% confidence). Increased risk was also quantified for spayed females (R = 4.2) and for all males (R = 3.0), compared to intact females (98.3% confidence). Heartworm cases in Cedar Ridge were clustered adjacent to areas of disrupted forest landscape, suggesting a focal mosquito vector originating from the disturbed area. In Tehama City, cases were evenly scattered throughout the town, suggesting highly dispersive mosquito vector populations. The spatial distribution of cases in Palo Cedro indicated a woodland-associated vector with limited dispersal ability. Dipetalonema reconditum cases were mostly clustered by neighborhood in Tehama City.