Table 3 Summary of reviewed articles
From: A systematic, realist review of zooprophylaxis for malaria control
References | Geographic location | Sample | Findings | Accounted for bed net? | Accounted for socio-economic factors? | Predominant mosquito species and characteristics (as reported by authors) | Animal-related variable | Effect |
|---|---|---|---|---|---|---|---|---|
Bogh et al. [11] | The Gambia | 102 pairs of children | An. gambiae s.s. and An. melas: no difference in HBI between cattle and non-cattle group. An. arabiensis: reduction of HBI by 30% in presence of cattle. No significant difference in sporozoite rate of all mosquito species in cattle compounds (0.97%) compared to non-cattle compounds (1.28%) | Yes | No | An. gambiae s.s. (72%), An. arabiensis (10%): relatively zoophilic, An. melas (18%): relatively zoophilic | Cattle present: children sleeping <20 m from at least one cow vs cattle absent: children sleeping >50 m from nearest cow (other livestock present but not considered) | Zooprophylaxis (An. arabiensis), none (An. gambiae s.s. and An. melas) |
Bogh et al. [10] | The Gambia | 102 pairs of children | No difference in parasite prevalence odds ratio between cattle and non-cattle group after adjusting for wealth. Adjusted OR 1.69 (CI 0.67–4.24), p = 0.26 | Yes | Yes | As above | As above | None |
Bouma and Rowland [16] | Pakistan | 2,042 slides examined over 2 years | Higher parasite prevalence in children from households owning cattle (15.2%) than children without (9.5%) Mantel–Haenszel χ2 = 9.6, p < 0.005. Mean parasite rates and prevalence of cattle keeping were positively correlated for seven villages (r = 0.79, p = 0.036) | No | No | Anopheles culicifaces: zoophilic, Anopheles stephensi: zoophilic, Anopheles subpictus: zoophilic | Cattle or water buffalo kept within the household compound | Zoopotentiation |
Bulterys et al. [40] | Zambia | 34 case households, 27 control households | Cattle ownership was associated with reduced odds of recurrent malaria infection (adjusted OR 0.19, CI 0.05–0.69). Households with the most cattle, goats, dogs, or cats had reduced odds of recurrent infection (adjusted OR 0.13, CI 0.03–0.56) | Yes | No | An. arabiensis: anthropophilic/opportunistic, An. funestus | Animal ownership (location not measured) | Zooprophylaxis |
Ghebreyesus et al. [23] | Ethiopia | 2,114 children (<10 years) | Animals sleeping indoors increased the incidence rate ratio for malaria infection (adjusted RR 1.92, CI 1.29–2.85). Cattle ownership was not associated with malaria infection (1–2 cows: aRR 0.75, CI 0.39–1.45; 3–4 cows: aRR 1.18, CI 0.65–2.14; ≥5 cows: aRR 1.18, CI 0.64–2.17) nor was sheep and goat ownership (1-4 sheep/goats: aRR 0.93, CI 0.58–1.50; ≥5 sheep goats: aRR 0.81, CI 0.54–1.22) | No | Yes | An. arabiensis | Cattle ownership, sheep and goat ownership, animals sleep inside house | Zoopotentiation for animals sleeping indoors. No effect for sheep/goat or cattle ownership. |
Habtewold et al. [12] | Ethiopia | 278 mosquitoes | No significant difference in proportion of mosquitoes feeding on humans and livestock for people sleeping with livestock indoors (site B) vs livestock housed separately (site A). Higher proportion of mosquitoes feeding on cattle (93.7%) compared to humans (3.1%) for people sleeping on elevated platforms (site C) above livestock (p < 0.05). Higher proportion of cattle feeding in site C (93.7%) vs sites A (42.7%) and B (54.7%) (p < 0.001) | No | No | An. arabiensis: moderately zoophilic | Humans sleep in traditional houses with cattle in separate enclosures (site A), humans sleep in houses with livestock sharing dwelling at night (site B), humans sleep in tree platforms above cattle (site C) | Zooprophylaxis |
Habtewold et al. [13] | Ethiopia | 18 study replications, total mosquito catch not reported | No effect of untreated ox on HBC for An. arabiensis but ox odour increased HBC (mean catch/person/night = 22 without cattle odour, 32 with, p < 0.05). For An. pharoensis HBC was significantly reduced in the presence of untreated ox (catch/person/night = 50 without and 26 with, p < 0.01) but increased in presence of cattle odour (catch/person/night = 6 without and 18 with, p < 0.001). CIs included but graph printing obscures visualization for most values | NA | NA | An. arabiensis: zoophilic, exophagic. Secondary vector: An. pharoensis | “Nearby” specific distance not reported | None (An. arabiensis), zooprophylaxis (An. pharoensis) |
Hadis et al. [36] | Ethiopia | 611 An. arabiensis mosquitoes | Mosquitoes collected from mixed human-livestock dwellings had significantly lower HBI (20.2%) than mosquitoes collected from human-only dwellings (91.5%) p < 0.001 | No | No | An. arabiensis | Human dwellings vs mixed human-cattle dwellings vs cattle sheds | Zooprophylaxis |
Hewitt et al.a [24] | Pakistan | 643 anopheline mosquitoes | HLC increased in presence of a cow (38%, CI 8–68%), and two goats (50%, CI 16–84%) | NA | NA | An. stephensi: zoophilic | A cow or two goats tethered 6 m from male mosquito collectors | Zoopotentiation |
Hiscox et al. [38] | Lao PDR | 879 anopheline mosquitoes | Cow ownership doubled the risk of anopheline house entry (IRR 2.32, CI 1.29–4.17, p = 0.005) | Yes | Yes | Anopheles philippinensis | Ownership of chickens, ducks, pigs, cows, or buffaloes, and keeping large animals (pigs, cows, buffaloes below the house) | Zoopotentiation for cow ownership but no effect of owning any other animals or keeping large animals below the house |
Iwashita et al. [33] | Kenya | 104 houses, 1,664 anopheline mosquitoes | An. arabiensis abundance increased by 10% with each additional goat/sheep tethered around the house. [Exp (β) = 1.10, β = 0.10, p = 0.02]. Odds of human blood feeding were decreased 0.99 times by each goat or sheep tethered within 500 m of the household [Exp (β) = 0.99, β = −0.01, p < 0.01] | Yes | No | An. arabiensis: zoophilic, exophagic, An. funestus s.s.: anthropophilic, endophagic, An. gambiae s.s. anthropophagic, endophagic | Cattle or goats/sheep kept within 20 m of house | None (An. gambiae s.s., An. funestus s.s), zoopotentiation (An. arabiensis) |
Lardeux et al.a [54] | Bolivia | 384 blood fed mosquitoes | Anopheles pseudopunctipennis preferred small ruminants (forage ratio 1.99, CI 1.80–2.19) to equids (1.95, CI 1.38–2.52) to humans (1.47, CI 1.25–1.69) to cows (1.15, CI 0.65–1.66) and avoided pigs (0.34, CI 0.20–0.48) and chickens (0.03, CI 0–0.75) | No | No | An. pseudopunctipennis: opportunistic | Various collection locations including outdoor traps and indoor resting collections | Zooprophylaxis |
Maia et al.a [21] | Ghana | 1,017 anopheline mosquitoes | Presence of cattle reduced the number of An. gambiae s.s. for HLC by 66% (p < 0.0001) but increased the density of Anopheles ziemanni (not statistically significant). Cattle presence did not influence the HLC number from 20 m away | NA | NA | An. gambiae s.s.: NA, An. ziemanni: zoophilic | Cattle inside 6 × 7 m experimental pen | Zooprophylaxis (An. gambiae s.s) |
Mala et al. [37] | Kenya | 20 households, 417 mosquitoes | Odds of An. arabiensis occurrence decreased in presence of animals (OR 0.4, p = 0.03) and odds decreased with increasing distance to animal shelters (OR = 0.88, p < 0.001) | No | No | An. arabiensis (66%), An. funestus (18%), An. pharoensis (15%) | Presence of animals, relative distance to animal sheds | Unclear |
Mutero et al. [8] | Kenya | 420 households | Low malaria prevalence in irrigated villages compared to non-irrigated villages (p < 0.05). Authors attribute this to preference for cattle feeding by An. arabiensis in the irrigated villages | No | No | An. arabiensis: zoophilic | Mean tropical livestock units per village | Zooprophylaxis |
Palsson et al. [35] | Guinea Bissau | 30 households | Presence of pigs indoors associated with increased mosquito abundance (χ2 = 17.63, p < 0.001) but the presence of goats was not (χ2 = 1.08, p < 0.30). Goats were relatively uncommon compared to pigs (relative prevalence of livestock not reported) | No | No | An. gambiae s.l. (An. gambiae s.s. most abundant) | Presence of pigs or goats inside the house | Zoopotentiation |
Temu et al. [41] | Mozambique | 8,338 children from 2,748 households | Pig keeping associated with increased odds of malaria infection (OR 3.2, CI 2.1–4.9) | Yes | Yes | An. gambiae s.s.: anthropophilic, An. funestus: anthropophilic | Children living in households with chickens, goats, sheep, cows, pigs | Zoopotentiation |
Tirados et al. [34] | Ethiopia | 63,194 mosquitoes | HLC caught significantly more mosquitoes (163 mosquitoes/trap/night) than CBT (26 mosquitoes/trap/night, F = 35.9, p < 0.001) outdoors in areas of high cattle: human ratio compared to areas of low cattle: human ratio (HLC = 3.1, CBT = 2.1, no significant difference reported) | NA | NA | An. arabiensis: anthropophilic, exophagic | Cattle: human ratio 0.6:1 vs 17:1. | Zoopotentiation |
Tirados et al.a [22] | Ethiopia | Not reported | Outdoor HLC of An. arabiensis was not affected by the presence of a surrounding cattle ring, while the presence of a surrounding cattle ring reduced the outdoor HLC for An. pharoensis by 44% (p < 0.05). Indoor HLC did not differ from outdoors for either vector species. The indoor HLC decreased by 49% (p < 0.01) in presence of cattle ring for An. arabiensis. The catch of An. arabiensis in HBT was 25 times greater than in CBT (p < 0.001) whereas, for An. pharoensis there was no significant difference. HBT and CBT catches were unaffected by a ring of cattle for either vector species | NA | NA | An. arabiensis: opportunistic, exophagic | Presence of a ring of 20 cattle surrounding the place where a person was (either outside or inside hut) | Zooprophylaxis |
Yamamoto et al. [43] | Burkina Faso | 117 cases, 221 controls (women and children <9 years) | In univariable analyses, keeping donkeys (OR 0.59, CI 0.34–1.01), rabbits (OR 0.52, CI 0.25–1.09), and pigs (0.26, CI 0.07–0.89) within the compound had a significantly protective effect at the p < 0.20 level. While no effect was found for cows (OR 0.84, CI 0.45–1.54), sheep (OR 0.84, CI 0.51–1.37), goats (OR 0.08, CI 0.60–1.93), or poultry (OR 1.14, CI 0.68–1.90). No difference between malaria cases and controls associated with animal ownership after adjusting for bed net use and level of education (odds ratio of multivariate analysis not reported) | Yes | Yesb | An. gambiae, An. funestus, An. arabiensis | Animals kept in compound | None |