Field trial of five repellent formulations against mosquitoes in Ahero, Kenya.
Arthropod repellents represent a first line of defense against biting arthropods. (1) The Department of Defense and other agencies are interested in repellent formulations to replace VV,VV-diethyl-3-methylbenzamide (deet) because of deet's chemical properties and safety concerns. Although deet is regarded as safe, registered with the Environmental Protection Agency, and has been in use over 5 decades, there have been incidences of serious adverse effects associated with the use of deet products, especially in infants and young children, (2) its chemical properties are damaging to some synthetic material and plastics, (3) and deet experienced a major public relations hit in the mid 1990s as it was suspected to have contributed to the so-called "Gulf War Syndrome." (4) Effective candidates, then, must be less caustic to the user, a nonplasticizer, and be at least as effective as deet. SS220 (1S, 2'S) 2-methylpiperidinyl-3-cyclohexene-1-carboxamide) and Bayrepel (picardin, 1-methyl-propyl 2-(2-hydroxyethryl)-1-piperidinecarboxylate) are considered such candidates.
The purpose of this study was to evaluate 2 formulations each of SS20 and Bayrepel against deet using volunteers acting as both treatment and their own control against Anopheles, Aedes, Coquillettidia, and especially Culex and Mansonia, the 2 most prevalent mosquito species in Kisumu, western Kenya. In order to confirm that the test leg repellent did not have a spatial affect on the other leg (the control leg), we included as one of the repellents a "null repellent", which was an application of no repellent at all. Volunteers treated with the null repellent, then, had an "application" of the "null repellent" on their treatment leg, while their other leg served as the control. In other words, there was no repellent on either leg.
Materials and Methods
This study was conducted at the homestead of a local inhabitant in the midst of a rice growing region near Kisumu, in the Kamagaga Village, Ahero Irrigation Scheme Sub-Location, Ombeyi Location, Miwani Division, Nyando District, Nyanza Province, Kenya, Africa (Lat -0.152098[degrees], Long 34.925649[degrees]) (Figures 1 and 2). There were ample species of mosquitoes due to the abundant breeding sites and recent seasonal rains in April and May 2004. Two field trials were conducted during May and July 2004.
The following 5 repellent formulations were used: (1 and 2) SS220 ((1S, 2'S) 2-methylpiperidinyl-3cyclohexene-1-carboxamide) formulated as a spray (20%) or a lotion (20%) by Avon (New York, NY); (3 and 4) 20% Bayrepel, (1-methyl propyl 2-(2-hydroxyethryl)-1-pi-peridinecarboxylate) formulated as a lotion by Avon or SC Johnson (Racine, Wisconsin); (5) 33% V,V diethyl-3-methylbenzamide (deet) formulated as a cream by 3M (St Paul, MN). For statistical design, a "null repellent" was introduced as the sixth repellent formulation, and consisted of no application at all. Note, the null repellent must not be confused with the control leg.
[FIGURE 1 OMITTED]
Male volunteers * from the local population were selected and screened for acceptance by a brief medical questionnaire, and blood pressure and pulse check. Each volunteer completed and was given a copy of a Kiswahili and English language human use protocol approved by the Kenya Medical Research Institute and the Walter Reed Army Institute of Research.
Twelve adult male volunteers participated in this study. Prior to repellent application, they washed both legs with clean water and allowed them to air dry. Clothing varied, as long as it prevented mosquito bites anywhere except the intentionally exposed skin area between the calf and ankle of each leg. The clothing usually included a light jacket, pants rolled to the knees, and locally purchased cotton work gloves and sport head nets which covered the entire head and neck. Since many volunteers wore open, slipper style shoes ("flip flops"), each foot was protected in a loosely fitting enclosure of mosquito bed net material which was gathered and taped around the volunteer's leg at the treatment line. As shown in Figure 3, the loose fitting ensured an air space around the foot (except for the sole), allowing olfactory cues to attract mosquitoes but not allowing them to feed. An area of 600 [cm.sup.2] was calculated and marked on each leg of each volunteer by averaging 3 equally-spaced circumference measurements between the lower knee and upper calf and dividing into 600 [cm.sup.2] in order to get the length. The resulting area was taped off to the pants at the top, and to the feet netting at the bottom. Each repellent was weighed for dose and applied evenly with a gloved-finger over the exposed skin on the treatment leg (Figure 4). To standardize procedure in the study design, the control leg was rubbed with a clean, dry, gloved finger. Similarly, the null repellent was also "applied," once again using a clean, dry, gloved finger.
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
The first volunteer's repellent was applied at 4 pm, followed by the other volunteers at 5 minute intervals. Evaluation was conducted on each of 6 nights every 2 hours starting at 6 pm (then 8 pm, 12 midnight, 2 AM, and 4 AM), each volunteer staggered at successive 5 minute intervals based on their application time. At the assigned time, the volunteer left the screened tent and walked the short distance (50 to 60 m) to the assigned test site. The test sites were separated from each other by at least 15 m. Each volunteer sat in an assigned location that was randomly determined, and a "collector/helper" sat opposite the volunteer. The collector/ helper was fully clothed and helped monitor the legs of the volunteer for landing mosquitoes in order to aspirate them before biting could occur (Figure 3). Volunteers remained in the test area for 20 minutes. Aspirated mosquitoes were expelled into lidded, pint-sized paper collection cartons. Each collection carton had a screen mesh in the lid for visibility, and a double dental dam portal on the side for the aspirating tube. The collected mosquitoes were immobilized by ether, counted, recorded, and transferred into vials in the field to be identified to species at a later date.
[FIGURE 4 OMITTED]
Totals at each time point were calculated for each repellent group and control. Percentage protection was defined as the number of bites received by a treatment group relative to that of the control ((control-treatment) control x 100).
Because there were no failures of 100% protection in the treatment group, we did not apply a statistical analysis, although had the protection been less than 100%, we would have performed an arcsine transformation before statistical analysis. We used a Student 2 tailed T test with groups having equal variance to compare the 3 control groups.
Results and Discussion
The goal of this study was to compare repellent activity of deet with potential candidate replacements SS220 and Bayrepel against mosquito species commonly found in Kisumu, Kenya.
Mosquito Populations Collected
In the first trial, a total of 3,137 mosquitoes were collected during these exposures. The main species were Mansonia uniformis (Theobald) (59.9%), Culex pipiens Linnaeus (18.1%), and Culex poicilipes (Theobald) (4.3%). Overall, 4 species of Aedes, 2 species of Anopheles, 3 species of Coquillettidia, 4 species of Culex, and 2 species of Mansonia were collected, as shown in Table 1. Figure 5 shows that Mansonia represented 76.6% of the total collected, and Culex represented 22.9%. It was surprising that few Anopheles were collected, as malaria transmission is a major problem in Kisumu. It is possible that the time of our study, 6 PM to 4 AM, did not coincide with either the main biting time of Anopheles, or the study site was not an optimal location where Anopheles may be found.
In the second trial, we focused only on the main species identified in the first trial. A total of 4,495 mosquitoes were collected, and of these, the most frequent were Culex pipiens (45.3%), Mansonia uniformis (42.2%) Culex poicilipes (10.7%) and Mansonia africana (Theobald) (1.6%). In this study, Culex represented 56.2% of the total collected, and Mansonia represented 43.8%. Therefore, the rates of collection varied from season to season, but the main species that predominated were similar, as were the collection rates.
Effect of Each Repellent on Mosquito Biting Rates
In the second trial, 10 of the 12 volunteers were used to test the 5 repellents, with one leg as the repellent, and the other leg as the repellent control. Two of the 12 volunteers were treated with null repellent on one leg and the other leg was the null repellent control.
We found that each repellent tested at the dose specified protected each treated leg 100% from mosquito bites, and there were no failures. Therefore, this trial did not distinguish whether SS220 and Bayrepel were less or more effective than deet. However, total protection lasted at least 8 hours after repellent application. It is clear that SS220 and Bayrepel were equally effective against the main mosquito species collected in Kisumu, Culex and Mansonia. A previous study indicated that piperidine compounds were less effective than deet in controlling Culex pipiens, (5) but the dose may have affected this outcome. Whether these repellents are as active against Anopheles, Aedes, or other genera would need confirmation in tests in areas where they are more prevalent than in Kisumu.
In a trial in Australia, (6) deet and Bayrepel were relatively much less active against Anopheles spp, where both protected volunteers for only up to one hour after application. However, deet and Bayrepel protected volunteers for up to 5 hours after application against Culex species, (6) comparable to our results in Kisumu. In another study in Burkina Faso, Bayrepel performed better than deet against Anopheles gambiae Giles, but did not repel Aedes. (7) However, Bayrepel and deet were effective against Aedes taeniorhynchus (Wiedemann).
A major problem with all such studies is the small sample size. Larger studies may be warranted to more effectively determine the activity of these repellents.
The method used in this study was to use each volunteer as his/her control, by treating one leg and leaving the other leg untreated. At each time, 10 volunteers were treated this way, and 2 volunteers were similarly treated using a null repellent (literally, no repellent at all). Surprisingly, the number of mosquitoes collected from the legs serving as the control for active repellents were statistically higher than either the null repellent (P = 0.001) or the null repellent controls (P = 0.003) (Table 2). However, there was no significant difference between the null treatment legs and the null treatment controls (P = 0.4) (Table 2).
There is little known about how exactly these repellents influence insect behavior. However, a recent study has suggested that they cause insects to move away from treated skin and bite only skin without the chemical. (8,9) This suggests that insects use a sense of smell to detect the chemicals and avoid biting where the repellent is coated. Our data suggests that insects were repelled from biting the treated skin and moved towards the untreated skin (controls), increasing the biting rates compared to the null repellents. We had no failures with the repellents in our study. However, this behavior effect would increase collections from controls and therefore underestimate the effectiveness of the repellent on the other leg if it is less than 100%. It seems, therefore, if a single volunteer is used as a repellent treatment and a no treatment control, it is essential that null treatment volunteers are also included to detect and correct such underestimation.
Finally, the wide spectrum of arthropods reacting to SS220 and Bayrepel against ticks, (10) biting midges (Leptoconops), (11) and the mosquitoes tested in this and similar studies, (12-16) suggest that these compounds may be valuable in controlling vector-borne diseases, in both military and civilian populations.
Studies such as reported here are valuable to better define appropriate evaluations of repellents under field conditions and to determine the mean protection periods. (12) Further investigation of the toxicity of repellents to mosquitoes may reveal more effective compounds or ones that may act synergistically, (13) as well as compounds that lack potential adverse effects of deet such as BioUD. (14) Recently, research measuring the toxicity of 8 repellents to female mosquitoes suggested that another piperidine compound, A13-31220, (15) may be more toxic at lower concentrations than deet. (16) Therefore, these single volunteer studies in Kenya and other regions may aid in the practical identification and acceptance of other, more effective repellent regimes.
The authors thank the following people, among many others, whose expertise, energy, and collegial support were absolutely integral to this project:
The Entomology staff at the US Army Medical Research Unit--Kenya, most notably, Maurice Agawo, Daniel Ngonga, and Francis Ngere from the Kisumu Field Station, and Christopher Oyaro, Entomology Department Senior Technician, Walter Reed Project and KEMRI.
MAJ Jason Richardson, USA, Armed Forces Research Institute of Medical Sciences, Bangkok, Thailand.
Pamela Hillis, Defense Logistics Agency.
Dr Desmond Foley, Walter Reed Biosystematics Unit.
Dr Gabriela Zollner and MAJ Jittawadee Murphy, USA, Walter Reed Army Institute of Research.
Also, we extend special appreciation to the Chief, Elders, villagers, and volunteers of Kamagaga and surrounding area. Asante sana!
(1.) Debboun M, Strickman DA, Klun JA. Repellents and the military: our first line of defense. J Am Mosq Control Assoc. 2005;21:4-6.
(2.) Osimitz TG, Grothaus RH. The present safety assessment of deet. J Am Mosq Control Assoc. 1995;11:214-218.
(3.) Katz TM, Miller JH, Hebert AA. Insect repellents: historical perspectives and new developments. J Am Acad Dermatol. 2008;58:865-811.
(4.) Qiu H, Won Jun H, McCall W. Pharmacokinetics, formulation, and safety of insect repellent N,N-diethyl-3-methylbenzamide (DEET): a review. J Am Mosq Control Assoc. 1998;14:12-21.
(5.) Coleman RE, Richards AL, Magnon GJ, Maxwell CS, Debboun M, Klein TA, Wirtz RA. Laboratory and field trials of four repellents with Culex pipiens (Diptera: Culicidae). J Med Entomol. 1994;31:11-22.
(6.) Frances SP, Waterson DG, Beebe NW, Cooper RD. Field evaluation of repellent formulations containing deet and picaridin against mosquitoes in Northern Territory, Australia. J Med Entomol. 2004;41:414411.
(7.) Costantini C, Badolo A, Ilboudo-Sanogo E. Field evaluation of the efficacy and persistence of insect repellents deet, IR3535, and KBR 3023 against Anopheles gambiae complex and other Afrotropical vector mosquitoes. Trans R Soc Trop Med Hyg. 2004;98:644-652.
(8.) Klun JA, Khrimian A, Debboun M. Repellent and deterrent effects of SS220, picaridin, and deet suppress human blood feeding by Aedes aegypti, Anopheles stephensi, and Phlebotomus papatasi. J Med Entomol. 2006;43:34-39.
(9.) Klun JA, Khrimian A, Rowton E, Kramer M, Debboun M. Biting deterrent activity of a deet analog, two DEPA analogs, and SS220 applied topically to human volunteers compared with deet against three species of blood-feeding flies. J Med Entomol. 2006;43:1248-1251.
(10.) Carrol J, Benenate J, Klun J, White C, Debboun M, Pound J, Dheranetra W. Twelve-hour duration of cream formulations of three repellents against Amblyomma americanum. Med Vet Entomol. 2008;22:144-151.
(11.) Perich MJ, Strickman D, Wirtz A, Stockwell SA, Glick JI, Burge R, Hunt G, Lawyer PG. Field evaluation of four repellents against Leptoconops americanus (Diptera: Ceratopogonidae) biting midges. J Med Entomol. 1995;32:306-309.
(12.) Rutledge LC, Gupta RK. Variation in the protection periods of repellents on individual human subjects: an analytical review. J Am Mosq Control Assoc. 1999;15 (3):348-355.
(13.) Debboun M, Strickman D, Klein TA, et al. Laboratory evaluation of AI3-31220, AI3-35165, CIC -4, and deet repellents against three species of mosquitoes. J Am Mosq Control Assoc. 1999;15 (3):342-347.
(14.) Witting-Bissinger BE, Stumpf CF, Donohue KV, Apperson CS, Roe RM. Novel arthropod repellent, BioUD, is an efficacious alternative to deet. J Med Entomol. 2008;45(5),891-898.
(15.) Xue RD, Arshad A, Crainich V, Barnard D. Oviposition deterrence and larvicidal activity of three formulations of piperidine repellent (AI3-37220) against field populations of Stegomyia albopicta. J Am Mosq Control Assoc. 2007;23(3),283-287.
(16.) Pridgeon JW, Bernier UR, Becnel JJ. Toxicity comparison of eight repellents against four species of female mosquitoes. J Am Mosq Control Assoc. 2009;25(2),168-113.
LTC Van Sherwood, MS, USA
Elizabeth Kioko, BS
Sichangi Kasili, MS
Philip Ngumbi, MS
Michael R. Hollingdale, PhD
* No female volunteers were used, as it was culturally unacceptable. The village chiefs and elders, whose endorsements were required for the trial to be performed at this site, deemed it unwise to have men and women together throughout the night.
LTC Sherwood is the Command Entomologist, Defense Logistics Agency, Fort Belvoir, Virginia. At the time this field study was conducted, he was Chief, Entomology Department, US Army Medical Research Unit--Kenya.
Elizabeth Kioko is a senior researcher with the Entomology Department, Walter Reed Project and Kenya Medical Research Institute, Kisumu, Kenya.
Sichangi Kasili is a senior researcher with the Entomology Section and Leishmaniasis Laboratory at the Centre for Biotechnology Research and Development, Kenya Medical Research Institute, Nairobi, Kenya.
Philip Ngumbi is Chief, Entomology Section and Leishmaniasis Laboratory at the Centre for Biotechnology Research and Development, Kenya Medical Research Institute, Nairobi, Kenya.
Dr Hollingdale is a biostatistical and research consultant at the Walter Reed Army Institute of Research, Silver Spring, Maryland.
Table 1. Distribution of mosquito species collected at the Kisumu area, Western Kenya, May 2004. Mosquito Species First Trial Second Trial Total % Total % Aedes (Albopictus) kennethi 1 < 0.1 NT * NT Aedes (Neomelanion) 1 < 0.1 NT NT luteolateralis Aedes (Neomelanion) 1 < 0.1 NT NT circumluteolus Aedes (Aedimorphus) cumminsii 1 < 0.1 NT NT Anopheles coustani 1 < 0.1 NT NT Anopheles funestus 1 < 0.1 NT NT Coquillettidia aurites 3 < 0.1 NT NT Coquillettidia fraseri 2 < 0.1 NT NT Coquillettidia fuscopennata 4 0.1 NT NT Culex annulioris 4 0.1 NT NT Culex pipiens 567 18.1 2,035 45.3 Culex poicilipes 136 4.3 483 10.7 Culex theileri 12 0.4 10 0.2 Mansonia africana 524 16.7 72 1.6 Mansonia uniformis 1,879 59.9 1,895 42.2 Grand Total 3,137 100.0 4,495 100.0 * Not tested Table 2: Total numbers of each mosquito collected on the repellent treated leg and its control untreated legs compared to null repellent controls at the Kisumu area, Western Kenya, May 2004. Leg Mean SD SE Repellent control 48.1 20.00 2.65 Null repellent treatment 29.9 12.10 3.82 Null repellent treatment control 34.4 12.67 3.65 Leg P Compared with Repellent Control Repellent control -- Null repellent treatment 0.007 Null repellent treatment control 0.003 Leg P Compared with Null Control Repellent control -- Null repellent treatment -- Null repellent treatment control 0.4 NOTE: The Student 2 tailed-T test where groups have equal variance was used to calculate whether more mosquitoes were collected on repellent control legs compared to null treatment and null treatment controls, or between null treatment and its null treatment control. Figure 5. Distribution of species mosquitoes collected during the first trial of the study in the Kisumu area, Western Kenya, May 2004. Mansonia uniformis 60% Mansonia africana 17% Culex pipiens 18% Culex poicilipes 4% All other species 1% Note: Table made from pie chart.