Insecticide-treated clothes for the control of vector-borne diseases: a review on effectiveness and safety

Abstract

Insecticide-treated clothing has been used for many years by the military and in recreational activities as personal protection against bites from a variety of arthropods including ticks, chigger mites, sandflies and mosquitoes. Permethrin is the most commonly used active ingredient, but others, including bifenthrin, deltamethrin, cyfluthrin, DEET (N,N-diethyl-3-methylbenz-amide) and KBR3023, have also been trialled. Treatment is usually carried out by home or factory dipping. However, new microencapsulation technologies which may prolong the activity of insecticides on clothing are now available and may help to overcome the inevitable reduction in efficacy over time that occurs as a result of washing, ultraviolet light exposure, and the normal wear and tear of the fabric. The aim of this article is to review the evidence base for the use of insecticide-treated clothing for protection against bites from arthropods and its effect on arthropod-borne pathogen transmission. Although some studies do demonstrate protection against pathogen transmission, there are surprisingly few, and the level of protection provided varies according to the disease and the type of study conducted. For example, insecticide-treated clothing has been reported to give between 0% and 75% protection against malaria and between 0% and 79% protection against leishmaniasis. Studies vary in the type of treatment used, the age group of participants, the geographical location of the study, and the pathogen transmission potential. This makes it difficult to compare and assess intervention trials. Overall, there is substantial evidence that insecticide-treated clothing can provide protection against arthropod bites. Bite protection evidence suggests that insecticide-treated clothing may be useful in the prevention of pathogen transmission, but further investigations are required to accurately demonstrate transmission reduction.

Introduction

There are many products and technologies designed to prevent bites from arthropods and to reduce the risk for transmission of vector-borne pathogens. Insecticide-treated bednets (ITNs) are probably among the most well-studied examples and good evidence of their efficacy against bites and of their significant effect on pathogen transmission has been reported [World Health Organization (WHO), 2011]. As well as the toxic mode of action of the insecticides used in ITNs, the nets themselves act as physical barriers to prevent biting and some of the active ingredients (AIs) may also have a repellent effect (Carter, 1989). However, bednets protect only against night-biting arthropods and therefore afford no protection against arthropods that bite during the day. This is of particular concern in relation to diseases such as dengue and yellow fever, the pathogens of which are vectored by day-biting mosquitoes (Hawley, 1988) of the genus Aedes (Stegomyia) (Diptera: Culicidae). For protection against those vectors, other technologies may be more appropriate and may include the use of topically applied repellents such as N,N,diethyl-m-toluamide (DEET), p-menthane-3,8-diol (PMD) and other non-topical repellent AIs including those that provide ‘spatial repellency’ (Fradin & Day, 2002). However, repellents are limited by variable user compliance. Short-term users, such as vacationers to endemic areas, have greater compliance than longterm or routine users such as outdoor workers (Achee et al., 2012) and this may greatly affect the effectiveness of repellents as a disease intervention.

Agricultural and wildlife groups, as well as commercial companies, use insecticide-treated clothing to protect workers and the public whilst in fields and forested areas (Breeden et al., 1982; Schreck et al., 1982; Vaughn & Meshnick, 2011). The military has used insecticide-treated uniforms for the protection of troops in vector-borne disease endemic areas for many years (Kitchen et al., 2009) and a recent mathematical modelling study demonstrated that insecticide-impregnated school uniforms have the potential to reduce dengue in children by 6–55% (Massad et al., 2013).

Despite the common use of insecticide-treated clothing, the evidence for its protective efficacy against arthropod bites and disease transmission is not clear or easily accessible. In this review, we examine the evidence base for the use of insecticide-treated clothing against insect bites and disease incidence. For the purpose of this review, the term ‘insecticide-treated clothing’ will refer to treated clothing effective against other arthropods including mites and ticks. We focus on the most commonly used insecticide, permethrin, to examine its safety profile and methods of impregnation. Although other insecticides, such as deltamethrin, bifenthrin and cyfluthrin, and AIs conventionally used as repellents, including DEET and KBR3023 (Table 1), have been used to treat materials, they are not discussed here.

Table 1. Studies investigating the different types of treatment and level of protection against biting arthropods

Active ingredient	Vector	Exposure time	Bite protection*, %	Knock-down, %	Mortality, %	Reference
Sprayed clothing
0.6 g/m^2 permethrin (Dragnet 100)	Aedes (Stegomyia) aegypti	3 min	6.1% (± 6.1)	NR	84.8% (± 8)	Frances et al. (2003)
0.6 g/m^2 permethrin (Dragnet 500)	Aedes (Stegomyia) aegypti	3 min	3.3% (± 3.3)	NR	97.0% (± 3.0)	Frances et al. (2003)
0.6 g/m^2 permethrin (Pergrin 500)	Aedes (Stegomyia) aegypti	3 min	6.1% (± 6.1)	NR	93.0% (± 3.5)	Frances et al. (2003)
0.6 g/m^2 permethrin (Dragnet 100)	Anopheles farauti	3 min	NR	100%	94.2% (± 2.5)	Frances et al. (2003)
0.6 g/m^2 permethrin (Dragnet 500)	Anopheles farauti	3 min	NR	100%	100%	Frances et al. (2003)
0.6 g/m^2 permethrin (Pergrin 500)	Anopheles farauti	3 min	NR	100%	100%	Frances et al. (2003)
0.5 g/m^2 permethrin (Imperator 25%)	Culex spp.	Dusk to dawn	69%	NR	27%	Rowland et al. (1999)
0.5 g/m^2 permethrin (Imperator 25%)	Anopheles nigerrimus	Dusk to dawn	65%	NR	4%	Rowland et al. (1999)
0.5 g/m^2 permethrin (Imperator 25%)	Anopheles stephensi	Dusk to dawn	31%	NR	0%	Rowland et al. (1999)
0.5 g/m^2 permethrin (Imperator 25%)	Anopheles subpictus	Dusk to dawn	37%	NR	34%	Rowland et al. (1999)
1.0 g/m^2 permethrin (Imperator 25%)	Culex spp.	Dusk to dawn	69%	NR	25%	Rowland et al. (1999)
1.0 g/m^2 permethrin (Imperator 25%)	Anopheles nigerrimus	Dusk to dawn	62%	NR	10%	Rowland et al. (1999)
1.0 g/m^2 permethrin (Imperator 25%)	Anopheles stephensi	Dusk to dawn	33%	NR	37%	Rowland et al. (1999)
1.0 g/m^2 permethrin (Imperator 25%)	Anopheles subpictus	Dusk to dawn	22%	NR	18%	Rowland et al. (1999)
2.0 g/m^2 permethrin (Imperator 25%)	Culex spp.	Dusk to dawn	76%	NR	44%	Rowland et al. (1999)
2.0 g/m^2 permethrin (Imperator 25%)	Anopheles nigerrimus	Dusk to dawn	43%	NR	40%	Rowland et al. (1999)
2.0 g/m^2 permethrin (Imperator 25%)	Anopheles stephensi	Dusk to dawn	58%	NR	39%	Rowland et al. (1999)
2.0 g/m^2 permethrin (Imperator 25%)	Anopheles subpictus	Dusk to dawn	0%	NR	51%	Rowland et al. (1999)
0.125 mg/cm^2 permethrin	Amblyomma americanum	1 h	100%	NR	100%	Schreck et al. (1982)
0.125 mg/cm^2 permethrin	Amblyomma americanum	15 min	98.5%	NR	NR	Evans et al. (1991)
0.125 mg/cm^2 permethrin	Ixodes dammini	15 min	100%	NR	NR	Evans et al. (1991)
0.125 mg/cm^2 permethrin	Total ticks	15 min	98%	NR	79%	Evans et al. (1991)
0.125 mg/cm^2 permethrin	Anopheles dirus	5 min	97%	99.4% (94.4–100%)	NR	Eamsila et al. (1994)
0.0025 g/m^2 bifenthrin (Biflex 10)	Aedes (Stegomyia) aegypti	3 min	3.3% (± 3.3)	NR	96.7% (± 3.3)	Frances et al. (2003)
0.0025 g/m^2 bifenthrin (Biflex 10)	Anopheles farauti	3 min	NR	100%	100%	Frances et al. (2003)
0.05 g/m^2 bifenthrin (Talstar 80)	Aedes (Stegomyia) aegypti	3 min	3.3% (± 3.3)	NR	96.7% (± 3.3)	Frances et al. (2003)
0.05 g/m^2 bifenthrin (Talstar 80)	Anopheles farauti	3 min	NR	100%	100%	Frances et al. (2003)
Hand-dipped
0.125 mg/cm^2 permethrin	Aedes (Stegomyia) albopictus	15 min	100%	60%	NR	Schreck & McGovern (1989)
10 g/m^2 KBR3020	Aedes (Stegomyia) aegypti	2 h	62.50% (95% CI 0.29–0.45)	NR	2.04% (95% CI 1.02–4.09)	Pennetier et al. (2010)
10 g/m^2 KBR3020 & 150 mg/m^2 pirimiphos-methyl (PM)	Aedes (Stegomyia) aegypti	2 h	76.56% (95% CI 0.18–0.30)	NR	2.48% (95% CI 1.33–4.68)	Pennetier et al. (2010)
0.125 mg/cm^2 permethrin	Glossina morsitans centralis (tsetse fly)	75 min	34.2%	NR	NR	Sholdt et al. (1989)
0.125 mg/cm^2 permethrin & 35% DEET (EDRF)†^,‡	Glossina morsitans centralis (tsetse fly)	75 min	90.7%	NR	NR	Sholdt et al. (1989)
0.125 mg/cm2 permethrin & 75% DEET (in EtOH)§	Glossina morsitans centralis (tsetse fly)	75 min	81.4%	NR	NR	Sholdt et al. (1989)
0.125 mg/cm^2 permethrin	Amblyomma americanum	15 min	97.5%	NR	NR	Evans et al. (1991)
0.125 mg/cm^2 permethrin	Ixodes dammini	15 min	100%	NR	NR	Evans et al. (1991)
0.125 mg/cm^2 permethrin	Total ticks	15 min	97%	NR	69%	Evans et al. (1991)
0.125 g/cm^2 permethrin	Trombicula spp. (chigger mite)	3 days	74.2%	NR	NR	Breeden et al. (1982)
2 g/ft^2 benzylbenzoate	Eutrombicula hirsti (chigger mite)	1 h	100%	NR	NR	Snyder (1946)
75% DEET	Phlebotomine sandflies	20 min, September 1980, January 1981	89% mean (range 68.8–98.1%)	NR	NR	Schreck et al. (1982)
0.125 g/m^2 permethrin	Phlebotomine sandflies	20 min, September 1980, January 1981	49% (range 6.8–90.5%)	NR	NR	Schreck et al. (1982)
0.125 g/m^2 permethrin	Phlebotomus papatasi	1, 3, 5, 7, 10 min	NR	8%, 26%, 56%, 56%, 84%	30%, 73%, 82%, 91%, 100%	Fryauff (1996)
0.125 g/m^2 permethrin	Culex pipiens	1, 3, 5, 7, 10 min	NR	49%, 70%, 75%, 85%, 98%	56%, 72%, 66%, 76%, 74%	Fryauff (1996)
2.5 mg/m^2 deltamethrin	Triatoma sordida	1, 2, 24 h (7-day mortality)	NR	0%, 0%, 20%	21.2%	Diotaiuti (2000)
2.5 mg/m^2 deltamethrin	Panstrongylus megistus	1, 2, 24 h (7-day mortality)	NR	60%, 100%, 100%	100%	Diotaiuti (2000)
2.5 mg/m^2 deltamethrin	Rhodnius neglectus	1, 2, 24 h (7-day mortality)	NR	100%, 100%, 100%	17.2%	Diotaiuti (2000)
2.5 mg/m^2 deltamethrin	Triatoma infestans	1, 2, 24 h (7-day mortality)	NR	30%, 40%, 100%	17.6%	Diotaiuti (2000)
5 mg/m^2 deltamethrin	Triatoma sordida	1, 2, 24 h (7-day mortality)	NR	0%, 0%, 10%	100%	Diotaiuti (2000)
5 mg/m^2 deltamethrin	Panstrongylus megistus	1, 2, 24 h (7-day mortality)	NR	45%, 100%, 100%	100%
5 mg/m^2 deltamethrin	Rhodnius neglectus	1, 2, 24 h (7-day mortality)	NR	100%, 100%, 100%	17.2%	Diotaiuti (2000)
5 mg/m^2 deltamethrin	Triatoma infestans	1, 2, 24 h (7-day mortality)	NR	30%, 60%, 100%	47%	Diotaiuti (2000)
Factory-dipped
0.125 g/m^2 permethrin	Aedes (Stegomyia) taeniorhyncus	9.5–10 h	99.9% (mean/day %)	NR	NR	Schreck & McGovern (1989)
0.125 g/m^2 permethrin	Aedes (Stegomyia) aegypti	2 h	56.25% (95% CI 0.35–0.55)	NR	11.66% (95% CI 8.43–16.30)	Pennetier et al. (2010)
0.125 mg/cm^2 permethrin	Culex sitiens	8 h	37.1%	NR	NR	Harbach et al. (1990)
0.125 mg/cm^2 permethrin	Aedes vigilax	8 h	43.1%	NR	NR	Harbach et al. (1990)
0.125 mg/cm^2 permethrin & 75% DEET (in EtOH)	Culex sitiens	8 h	72.9%	NR	NR	Harbach et al. (1990)
0.125 mg/cm^2 permethrin & 75% DEET (in EtOH)	Aedes vigilax	8 h	83.4%	NR	NR	Harbach et al. (1990)
0.125 mg/cm^2 permethrin & 35% DEET (EDRF)	Culex sitiens	8 h	78.8%	NR	NR	Harbach et al. (1990)
0.125 mg/cm^2 permethrin & 35% DEET (EDRF)	Aedes vigilax	8 h	93.5%	NR	NR	Harbach et al. (1990)
Permethrin	Anopheles spp.	6 h	48.4% (95% CI 0.63)	NR	11.3% (95% CI 1.66)	Deparis et al. (2004)
Permethrin & 50% DEET	Anopheles spp.	6 h	44.6% (95% CI 0.93)	NR	11.3% (95% CI 1.27)	Deparis et al. (2004)
0.125 mg/cm^2 permethrin cotton cloth	Pediculus humanus L. (body lice) (field strain)	10, 15, 30, 45, 60, 75 min, 12 h	NR	0%, 23.3%, 48.1%, 75.7%, 90.7%, 96.7%, 100%, 100%, 100%	NR	Sholdt et al. (1989)
0.125 mg/cm^2 permethrin NYCO blend cloth	Pediculus humanus L. (body lice) (field strain)	10, 15, 30, 45, 60, 75 min, 12 h	NR	18.9%, 44.4%, 82.0%, 87.9%, 92.5%, 100%, 100%	NR	Sholdt et al. (1989)
0.125 mg/cm^2 permethrin NYCO blend cloth	Pediculus humanus L. (body lice) (field strain)	15, 30, 60 s	NR	(68%, 30%, 32%)¶ (64%, 40%, 50%)**     (100%, 57%, 93%)††    (100%, 100%, 100%)‡‡	NR	Sholdt et al. (1989)
0.125 mg/cm^2 permethrin cotton cloth	Pediculus humanus L. (body lice) (laboratory strain)	15, 30, 45, 60, 75, 90, 105, 120, 135 min, 24 h	NR	80%, 99%, 96%, 100%, 100%, 100%, 100%, 100%, 100%, 100%	NR	Sholdt et al. (1989)
0.125 mg/cm^2 permethrin NYCO blend cloth	Pediculus humanus L. (body lice) (laboratory strain)	15, 30, 45, 60, 75, 90, 105, 120, 135 min, 24 h	NR	72%, 96%, 100%, 100%, 100%, 100%, 100%, 100%, 100%, 100%	NR	Sholdt et al. (1989)
Permethrin§§	Ixodes ricinus	27 weeks	99%¶¶    , 0.01 incidence rate ratio (95% CI 0.001–0.11)	NR	NR	Vaughn & Meshnick (2011)
Polymer-coated
1200 mg/m^2 permethrin	Ixodes ricinus	36 h	95.5%	NR	NR	Faulde et al. (2009)
Hand-applied
33.25% DEET (EDRF)	Amblyomma americanum	15 min	60.4%	NR	NR	Evans et al. (1991)
33.25% DEET (EDRF)	Dermacentor variabilis	15 min	50%	NR	NR	Evans et al. (1991)
33.25% DEET (EDRF)	Ixodes dammini	15 min	10%	NR	NR	Evans et al. (1991)
33.25% DEET (EDRF)	Total ticks	15 min	59.8%	NR	NR	Evans et al. (1991)

• * Bite protection: values in bold represent the percentage blood-fed after exposure; values in standard font represent the reduction in the percentage of bites.
• † Repellent applied to the skin while concurrently wearing treated clothing.
• ‡ EDRF, extended duration repellent formulation or controlled release formulation.
• § EtOH, ethanol.
• ¶ 15 s, 30 s, 60 s exposure at 0.5 h knock-down.
• ** 15 s, 30 s, 60 s exposure at 1.0 h knock-down.
• †† 15 s, 30 s, 60 s exposure at 6.0 h knock-down.
• ‡‡ 15 s, 30 s, 60 s exposure at 12.0 h knock-down.
• §§ No concentration of permethrin given; clothing provided by Insect Shield combining factory-based coating technology with a proprietary formulation of permethrin.
• ¶¶ Rate of tick bites acquired during work hours was reduced by 99%
• NR, not reported; 95% CI, 95% confidence interval.

Materials and methods

We used the following search engines: PubMed; BIOISIS Citation Index; Medline; Google Scholar, and the Armed Forces Pest Management Board Literature Retrieval System. We used the following terms for our search: ‘insecticide-treated clothes’; ‘insecticide-treated materials’; ‘repellent-treated materials’; ‘permethrin’; ‘DEET’; ‘polymer coating and textiles’; ‘microencapsulation for treated clothing’; ‘disease intervention using insecticide-treated clothing’; ‘pyrethroids’; ‘permethrin safety’; ‘absorption rates of permethrin’; ‘DEET safety’; ‘biting times’; ‘guidelines for testing insecticides’, and ‘spatial repellents’. All articles were located between October 2011 and May 2013.

Results

Active ingredient

The most commonly used AI is the pyrethroid permethrin (Table 1). It is thought to act as both a repellent and insecticide and produces Type I responses, which elicit repeated firing of neurons resulting in the blocking of neural messages (Sonderland et al., 2002). The ‘repellent effect’ is defined by increased movement displayed by arthropods and specifically detachment in ticks once they make contact with the material (Faulde & Uedelhoven, 2006). Permethrin has been shown to be more effective at deterring ticks than DEET, which is one of the most effective arthropod repellents available (Brown & Herbert, 1997). Not only does permethrin cause ‘hot feet’, whereby insects appear to be standing on a hot surface and repeatedly lift their legs to avoid contact with material, but permethrin also knocks down and kills biting arthropods through its insecticidal properties (Rossbach et al., 2010). Permethrin has displayed high bite protection against a wide range of biting arthropods including ticks (97%) (Evans et al., 1991) and mites (74%) (Breeden et al., 1982). Testing with mosquitoes has shown effectiveness that ranges from high bite protection against Aedes albopictus (Stegomyia albopicta) (100%) (Schreck & McGovern, 1989) to low protection against Culex sitiens (Diptera: Culicidae) (37%) (Harbach et al., 1990) (Table 1). Although permethrin is a potent insecticide, it has low toxicity in mammals and is used widely in nuisance and disease vector pest control treatments for humans and cattle (WHO, 2005; Appel et al., 2008).

Permethrin safety

Despite permethrin's effectiveness and safety, a small amount of the insecticide is still absorbed through dermal contact and the pyrethroid is listed as a class C carcinogen (Snodgrass, 1992; Appel et al., 2008). This potential risk requires that all contact with permethrin be monitored and the pyrethroid's metabolism through the body measured. The U.S. Environmental Protection Agency (EPA) has set up regulations to evaluate all forms of absorption, ingestion and uptake of permethrin into the body, and only allows those that fall below the maximum daily exposure (U.S. EPA, 2011). Measuring daily uptake in individuals using permethrin-impregnated clothing on a longterm basis, such as military and domestic uniformed personnel, is an important safety consideration (Rossbach et al., 2010). The EPA has calculated non-cancer and cancer risks for military and non-military workers who wear permethrin-impregnated clothing. The EPA estimated the clothing (0.125 mg AI/cm²) would be worn for 250 days per year for 10–35 years. Non-cancer risk (for illnesses other than cancer) is determined by calculating the margin of error (MoE) and only MoEs of > 100 are considered below any level of risk. Cancer risk is based on the likelihood of one to three people in one million (1–3 × 10⁻⁶) developing cancer as a negligible risk. These levels are based on dermal absorption rates, body weight, number of days per year worn, the surface area of skin in contact with the clothing, and the concentration of the AI. Studies by the EPA estimate MoEs of 6700 and 26 000 in military and non-military use, respectively. Cancer risk was calculated at 1.2 × 10⁻⁶ for military and 3.6 × 10⁻⁶ for non-military workers, both of which fall under the EPA level of concern (U.S. EPA & Prevention, 2009). Although some studies have raised concern over potential neurological and cellular negative effects seen when permethrin is used in combination with DEET in rats, the doses used were 0.13 mg/kg/day and the AI was applied directly to rats (Abu-Qare & Abou-Donia, 2001; Abdel-Rahman et al., 2004). This dose is well above the 0.00068 mg/kg/day allowed for humans (Snodgrass, 1992). The EPA has completed testing in pre- and post-natal studies along with developmental toxicity and has found no evidence that permethrin poses extra risk to children and infants (U.S. EPA & Prevention, 2009). Multiple studies have shown that wearers of insecticide-treated clothing have absorption rates that fall five-fold below the regulations established by the WHO and EPA (Snodgrass, 1992; Appel et al., 2008; Rossbach et al., 2010). However, the methods of impregnation do show differing levels of exposure risk. Dipped uniforms carry a greater exposure risk than commercial aerosol spray applications (Snodgrass, 1992). Polymer coating techniques appear to be associated with the lowest absorption rate but remain capable of impregnation with greater doses of permethrin (Faulde & Uedelhoven, 2006). However, careful monitoring of the products available and the concentrations of permethrin in them is required to maintain these acceptable daily rates. There is some indication that the absorption of permethrin into the skin is greater in warm climates (Snodgrass, 1992; Appel et al., 2008; Rossbach et al., 2010); however, robust scientific evidence for this is lacking. Its high efficacy and the wide range of arthropods against which permethrin protects, along with its low toxicity, make it an ideal substance for the impregnation of clothing.

The treatment process

Impregnation of clothing with insecticides and repellents is achieved by binding the AI to the fabric, utilizing four different techniques: absorption; incorporation; polymer coating, and microencapsulation. All techniques require a fixative, a polymer material also known as a binder, to hold the AI to the fabric throughout wear and washing (Marinkovic et al., 2006). The first method, absorption, allows the AI to bind to material by either spraying or dipping fabric in the insecticide along with a chemical to facilitate the binding process (Schreck et al., 1982). The incorporation method, which is mainly used to treat carpets (Williams et al., 2003), will not be discussed further in this review. The polymer-coating method employs a layer of polymers, within which the AI is bound, that is coated over the fibre surfaces (Faulde & Uedelhoven, 2006). The final technique is microencapsulation, which is similar to polymer coating, but encases the AI in a capsule that is mixed into a binding solution. The material is then run through a bath of the AI solution, allowing for a thin layer of polymer which gives the advantage of binding into the fibres (Marinkovic et al., 2006; Appel et al., 2008).

The absorption method can be used in both individual use and large-scale factory production. Home spraying or dipping kits are affordable, available worldwide, and represent an easy-to-use option for treating clothing with permethrin. However, because of variation in personal application, the coverage may be inconsistent across the article of clothing and as a result some areas may be vulnerable to biting arthropods (Schreck et al., 1982). Home-dipping kits do not provide long protection times and reapplication after five washes is often recommended (Kimani et al., 2006). Alternatively, factory spraying or dipping, which occurs on a large scale in a standardized process, gives a more consistent application to the whole article of clothing. Companies that produce clothing using this method claim that efficacy against biting arthropods lasts up to 70 washes (Vaughn & Meshnick, 2011). Although this type of impregnation is also readily available to the average consumer, it is more expensive than that obtained using home kits.

The use of polymers to bind permethrin to fabric has been found to provide longer-lasting protection than other methods (Faulde & Uedelhoven, 2006; Faulde et al., 2009). The polymer-coating method is performed on a large factory-based scale in which the AI is added to the polymer coating before the pre-tailored fabric is treated. This method is more expensive than dipping methods, but has the benefit of lasting up to 100 washes (Faulde & Uedelhoven, 2006) and allows for the addition of more insecticide with lower absorption rates into the skin (Faulde et al., 2012). Another benefit is the reduction in the environmental impact of washing clothing treated with insecticide in this manner as there is less run-off in the washing liquid (Snodgrass, 1992; Rossbach et al., 2010).

Microencapsulation of permethrin involves enclosing permethrin in a shell of a low-molecular pre-polymer or monomer, and allows for a specified rate of release, lower absorption by humans and longer stability of the insecticide (Marinkovic et al., 2006). After the fabric has been soaked in microencapsulated permethrin, the material is rolled through a pressing and heating section (Marinkovic et al., 2006). In microencapsulation, the fabric is treated with an ultra-thin layer of polymer that does not just rest on the outside of the fabric, but is thin enough to work its way around individual fibres (Marinkovic et al., 2006). This treatment could allow for longer-lasting release and stability of the insecticide without changing the feel, thickness and appearance of the clothing. Changes to feel, thickness and appearance may have an impact on compliance, especially in hotter climates in which heavier, thicker clothing is not worn. However, no efficacy studies on this type of treatment have been published.

Evidence for the efficacy of impregnated clothing

The efficacy of insecticide-treated fabrics is usually measured by recording how efficiently the product provides knock-down and killing of biting arthropods. However, many products claim protection against biting and many studies include biting protection as a measure of efficacy. Bite protection is important to quantify as this has the potential to reduce or prevent the transmission of vector-borne pathogens. However, bite protection is a reflection of the repellency of the AI and does not consider any of the longterm effects of mortality or the reduction of vector populations. These measures of efficacy can be investigated by performing cone tests recommended by the WHO Pesticides Evaluation Scheme (WHOPES, 2005) and arm-in-cage tests on a material after weathering and washing is simulated (Gupta et al., 1990). There are currently no WHOPES guidelines for the testing of insecticide-treated clothing. Several investigations have shown varying degrees of efficacy against a number of arthropods; these studies are summarized in Table 1.

Bite protection

Although insecticide-treated clothing can provide some protection against biting, few studies have shown 100% protection against bites using insecticide-treated clothing alone (Table 1). A study reported in 1989 found that bites were mainly distributed on exposed skin when subjects were wearing treated uniforms. The opposite was found when subjects only wore repellent and 100% of the bites occurred through the untreated clothing (Schreck & McGovern, 1989). However, protection was increased to nearly 100% when subjects wore insecticide-treated clothing and a topically applied repellent (Schreck et al., 1982, 1984; Schreck & McGovern, 1989; Sholdt et al., 1989; Pennetier et al., 2010). This combination of repellent and insecticide-treated clothing is recommended by the U.S. Department of Defense for the protection of military personnel against biting arthropods (Armed Forces Pest Management Board (AFPM), 2009) and has been shown to give 99.9% protection against biting arthropods over a 9-h period (Schreck et al., 1984). However, despite the success of this combination approach, biting on the head has been shown to increase when both treatments are used (Sholdt et al., 1989).

One factor to consider when interpreting the results of bite protection studies is that different fabric treatment processes can result in different levels of bite protection as a result of variability in the absorption and coverage that occurs in the process of applying the insecticide or repellent to the surface, or soaking it into the material. Differences in the duration and type of assay used to assess the treated material will also provide variable results, making direct comparisons difficult (Table 1). Future comparative studies, using standard guidelines, of types of treatment and assays may help to resolve these issues.

Reduction in insect populations: effect of knock-down and mortality

Knock-down and mortality caused by insecticide-treated materials will inevitably protect the individual from bites. The ability to knock down and kill also has the potential to protect at the community level by reducing the population of arthropods in a given area (Seng et al., 2008). This longlasting community effect has been demonstrated in other studies using insecticide-treated nets and curtains (Kroeger et al., 2006; Seng et al., 2008) and has succeeded in reducing insect populations up to 100 m away from the intervention (Lenhart et al., 2008). For example, Kimani et al. (2006) collected mosquitoes from the homes of individuals who wore insecticide-treated clothing and compared them with collections from the homes of individuals who wore untreated clothing. Collections were carried out in a baseline survey and then once per month over 3 months. A significant reduction in total mosquito density was found between treated and untreated homes, whereby treated homes showed a reduction in total density, engorged mosquito density, female mosquito density, and Anopheles (Diptera: Culicidae) density (Kimani et al., 2006). Rowland et al. (1999) conducted an efficacy test on mortality using individuals wearing treated ‘chaddars’, which are large pieces of cloth used as body covers during both the day and night by women and children, under untreated nets. Human- and cow-baited traps were used in order to attract host-seeking mosquitoes. The trapped mosquitoes were then collected and released into an untreated net, under which sat an individual protected with a treated chaddar. A significant reduction in mosquito numbers inside the untreated net with the individual wearing a treated chaddar was reported for Culex spp., Anopheles stephensi and Anopheles nigerrimus (Rowland et al., 1999). These two studies demonstrate the potential for insecticide-treated clothing to reduce populations of mosquitoes when they come into contact with the material. The results are promising and the community effects should be investigated further with longer follow-up periods and wider collection radii.

In a similar way to the bite protection studies, a lack of standardized methods is apparent in studies that have examined effects of impregnated materials on insect populations (Table 1). Duration of exposure, the period of time used to measure knock-down and mortality, the definition of bite protection, and the concentration of the AI should all be considered, but many of these factors are omitted from the published studies. The use of standard methods in all future studies would help us to better assess the efficacy of insecticide-treated clothing.

Intervention trials: effects of insecticide-treated clothing on disease

Several intervention trials have been performed using insecticide-treated materials; however, results range from no demonstrated reduction in pathogen incidence to reductions of up to 70% for malaria Eamsila et al., 1994; Asilian et al., 2003) (Table 2).

Table 2. Study characteristics of trials of pesticide-treated clothing against vector-borne pathogen transmission

Location (authors)	Trial characteristics
Afghanistan (Rowland et al., 1999)	Trial type: randomized controlled trial, malaria intervention
	Study population: 3950 refugees (510 families) from the Adizai refugee settlement. 825 individuals from 102 households were randomly selected. Mean age in the permethrin-treated group was 17 years; mean age in the placebo group was 19 years. Diagnosis was verified by microscopy at the local health centre
	Blinding: field staff and study participants were blinded to allocation
	Intervention: 0.1 mg/cm^2 permethrin-treated chaddars,* patoos and top sheets. Chaddars are pieces of material often used for protection against the sun, and biting insects. At night they are used for warmth. During the cold months patoos are used as they are made of thicker material
	Other data recorded: age; gender; social acceptance of insecticide treatment and compliance; clothing washing habits, and insecticidal efficacy against local mosquito population
	Outcome: the effect of treated chaddars, patoos and top sheets on the reduction of Plasmodium falciparum (P < 0.001) was significant, but only borderline for Plasmodium vivax malaria (P = 0.069). Adjusted odds ratios (95% confidence intervals): P. falciparum 0.51 (95% CI 0.30–0.86), and P. vivax 0.70 (95% CI 0.43–1.13). Highest incidence risk group 0–10 years and lowest > 20 years identified
	Length of trial: 16 weeks
Thailand (Eamsila et al., 1994)	Trial type: randomized controlled trial, malaria intervention
	Study population: 403 Thai soldiers aged 18–40 years from malaria-free areas with little or no acquired immunity. Mid-trial first group of soldiers were sent home and a new group of 260 troops replaced them. Three locations were used along the Thai–Cambodian border. Diagnosis was verified through microscopy
	Intervention: 0.125 mg/cm^2 permethrin-treated uniforms; untreated bednets were also available to volunteers
	Outcome: the intervention had no significant effect on either P. falciparum or P. vivax malaria incidence over the 13 months
	Length of trial: 13 months
Kenya (Kimani et al., 2006)	Trial type: randomized controlled community trial with a treatment and comparison arm, malaria intervention
	Study population: 198 Somali refugees in Dadaab refugee camp. Multistage cluster sampling was used to select participants. Diagnosis was verified through microscopy
	Blinding: double blinding of participants and laboratory staff
	Intervention: personal clothing of the participants was permethrin-dipped at a concentration of 15 mL of permethrin : 4000 mL of water. Repeated every 3 weeks
	Other data recorded: baseline blood samples; mosquito density; sleeping habits; malaria control practices; age; gender; clothes washing habits, and possession of treated bednets
	Outcome: owning a bednet reduced the odds ratio of getting malaria to 0.30. The odds ratio post-intervention was reduced by 69% (OR = 0.31, P < 0.001). Found intervention protective of all age groups except 0–5 years and 25–49 years
	Length of trial: 3 months
Colombia (Soto et al., 1995)	Trial type: randomized controlled trial, malaria and leishmaniasis intervention
	Study population: 172 Colombian soldiers for malaria trial and 286 Colombian soldiers for leishmaniasis trial. Soldiers were sent to malaria and leishmaniasis endemic areas, and had no previous or current signs of infection. Diagnosis for malaria was verified through microscopy; diagnosis for leishmaniasis was verified through examination of lesions followed by microscopy and electrophoresis
	Blinding: field staff and study participants were blinded to allocation
	Intervention: 600–712 mg/m^2 permethrin-treated uniforms. Instructed to wear the uniforms day and night
	Other data recorded: compliance; lesion placement (used as an indicator of compliance), and use of other repellents
	Adverse effects: two of 229 participants reported irritation and pruritis
	Outcome: a significant reduction in malaria of 79% (P = 0.015) was reported. Although a lower reduction of 75% in leishmaniasis was seen, it was determined to be more significant (P = 0.002)
	Length of trial: malaria study involved 3–5 weeks in endemic area and a 4-week follow-up (mean total length: 8.2 weeks). Leishmaniasis study was 6–8 weeks in endemic area followed by 12 week follow-up (mean total length: 18.6 weeks)
Iran (Asilian et al., 2003)	Trial type: randomized controlled trial, cutaneous leishmaniasis intervention
	Study population: 272 Iranian soldiers between the ages of 19 and 24 years with no history of cutaneous leishmaniasis. 134 soldiers in treatment group and 138 in control. Located in Isfahan, and endemic area. Diagnosis was verified by microscopy
	Blinding: field staff and study participants were blinded to allocation
	Intervention: 850 mg/m^2 permethrin-treated uniforms; uniforms covered whole body except head, neck, hands, and feet. They were instructed to wear them day and night. Told not to wear any insect repellents
	Other data recorded: compliance, and any adverse reactions to the clothing
	Adverse effects: none
	Outcome: 4.4% in treatment group were infected, and 6.5% in the control group were infected with leishmaniasis. No significant reduction in leishmaniasis transmission was provided by wearing the permethrin-treated uniforms (P < 0.05)
	Length of trial: 9 months
Afghanistan (Faulde et al., 2009)	Trial type: year-long programme with synergistic personal protection techniqes and disease control measures, zoonotic cutaneous leishmaniasis intervention
	Study population: German field camp for up to 1200 soldiers in Mazar-e Sharif airport, northern Afghanistan. Diagnosis was verified by polymerase chain reaction
	Intervention: theoretical (not measured after impregnation) concentration of 1300 mg/m^2 permethrin polymer-coated uniforms. Theoretical 25 mg/m^2 deltamethrin-treated bednets and curtains were completed annually. 0.005% bromadialon poison-baited traps were used for rodent control. Residual heat fogging of rodent burrows and sandfly breeding sites, and non-residual insecticide fogging near living and work areas. Camp sanitation procedures were implemented, including a high stone wall around the camp; soil compaction and stone paving; removal of upper layer of earth; compaction of earth in the surrounding 100 m area; and manual removal of all camp vegetation. Personnel were educated on the threat and transmission of vector-borne diseases, the proper timing and use of the treated uniforms, peak sandfly activity times, use of repellents, and the safest areas to spend leisure and sport time
	Other data recorded: surveillance of sandfly and rodent populations throughout the trial
	Outcome: infection rates of leishmaniasis were reduced. The previous year (2005) was used for baseline data for infection rates. 2005 had 17.5% (14 cases/80 persons) infection rate; 2006 had 0.087% (one case/1150 persons); and 2007 had 0% (no cases). Quantified infection rates were 0.058 in 2005, 0.0000055 in 2006, and 0.0 in 2007
	Length of trial: 1 year

• * A 3.5-m² piece of material used to clothe and protect the wearer; worn both day and night.

One study investigated the use of permethrin-treated clothing and bedding and demonstrated a 69% reduction in malaria transmission (Kimani et al., 2006). Similar results (64% reduction in malaria) were found in children aged 0–10 years in a study completed by Rowland et al. (1999), who implemented the use of treated chaddars, patoos (blankets made of thin wool) and top sheets in an Afghan refugee camp in northwest Pakistan. Additionally, Kimani et al. (2006) found reductions in the odds ratios of contracting malaria in subjects aged 5–24 years and in those aged > 50 years. In both trials, the clothing and materials treated were the participants' personal items, rather than new clothing. The utilizing of existing clothing and habits promotes stronger compliance and greater acceptability. The greater disease risk of the two age groups in which the incidence of malaria was reduced following the intervention in this study suggests that treated clothing may be a promising disease intervention method.

A large integrated vector management trial in northern Afghanistan used permethrin-treated clothing in an intervention against leishmaniasis (Faulde et al., 2009). The trial included the use of treated clothing, skin repellents, bednets and curtains, vector monitoring, habitat cleansing and health education programmes. Although the specific effect that treated clothing had on transmission rates is not clear, a pattern of decline was seen over the 3-year trial. Leishmaniasis infection rates fell from 17.5% in 2005 to 0% in 2007 (Faulde et al., 2009). This integrated method gave the greatest reduction in disease transmission in comparison with the other intervention studies reviewed. However, the study was completed in a manner that examined only reduction from the trial start date to its end date and had no control group, which makes it difficult to definitively state that the interventions caused the reduction in disease transmission.

Intervention trials have mainly focused on two diseases: malaria and leishmaniasis. This may reflect the number of studies completed by the military and the relevance of malaria and leishmaniasis to soldier well-being. In one study, which investigated the effect on scrub typhus transmission (Welt, 1947), four and a half battalions were divided into three test groups. One group received no miticide, the second group wore uniforms sprayed with miticide, and the third group wore uniforms dipped in miticide. The study found a high reduction of scrub typhus in groups 2 and 3 in comparison with group 1; however, whereas group 3 slept on cots, groups 1 and 2 slept on the ground. Not sleeping on the ground is a method of reducing scrub typhus and thus it is difficult to determine whether the reduction reflected the treatment of clothing or the sleeping location. Group 3 also did not visit the same locations as groups 1 and 2 (Welt, 1947). Because of these variations among the groups, no firm conclusion as to whether the clothing was responsible for the reduction in scrub typhus can be made.

Further difficulty arises in the comparison of trials as a result of variation among the methods used for each trial. Multiple treatments (e.g. spraying, hand-dipping and factory-dipping), duration and follow-up of the trial, the education of participants about the intervention, and the sample size, are all factors that vary across the trials that have been completed to date (Table 2).

Factors affecting efficacy

The acceptability of and compliance with treated clothing represent major challenges to the successful completion of an intervention trial. Although not all of the trials examined here elaborated on factors of acceptability, some, including those by Rowland et al. (1999), who reported participants' high and ‘enthusiastic’ acceptance of treated clothing, and Kimani et al. (2006) described a beneficial reduction in the number of mosquitoes and protection from mosquito bites. Eamsila et al. (1994) discussed a general feeling of compliance, but did not monitor participants' use of the intervention materials and listed low compliance and acceptability as potential limitations to the study.

Although the intervention trials make no direct link between education and compliance, Faulde et al. (2009) describe an emphasis on education in the use of intervention materials and risks associated with the intervention. A study conducted in Colombia, which did not monitor compliance, found that bites and lesions in all but one soldier were located on areas not covered by the treated clothing, indicating the compliance of trial participants (Soto et al., 1995). A third study, which used treated uniforms as an intervention against cutaneous leishmaniasis for soldiers in Iran made no mention of compliance or the location of lesions (Asilian et al., 2003) and therefore whether or not there is a correlation between participant acceptance and the success of the trial is unknown.

The efficacy of the AI is also affected by exposure to sunlight (ultraviolet light), the wear of the fabric, the washing of the material (Atieli et al., 2010) and the type of fabric (Amalraj et al., 1996). These factors all impact on how efficiently the insecticide remains active in the material. Additionally, the type of fabric affects the original absorption and binding rate of the permethrin to the material (Amalraj et al., 1996). General wear of insecticide-treated clothing causes deterioration of the fabric and permethrin, sunlight breaks down permethrin, and washing rinses out permethrin (Deparis et al., 2004; Faulde et al., 2012). For most treated fabrics such as bednets and curtains, these factors can be easily monitored and standardized. Bednet exposure to sunlight can be limited, wear is localized to specific regions on the net and bednets are washed less frequently than everyday clothing (Binka & Adongo, 1997; Rowland et al., 1999; WHO, 2007). Clothing, by contrast, is utilized in a more uncontrolled manner: sunlight is unavoidable in practical terms; wear can be assumed to occur but may vary between persons, and clothing is likely to be washed significantly more frequently than bednets. The breakdown, deterioration and loss of permethrin will affect the pesticide's ability to knock down and kill biting arthropods.

The effect of weathering on treated clothing has also been examined through the simulation of environmental factors on material. For example, Gupta et al. (1989) showed that most permethrin was lost in the first week of simulated weathering. The weathering involved a day/night light cycle along with changes of temperature and humidity to simulate the cycles of outdoor wear, such as from tropical to arid environments, using a Weather-ometer machine (Gupta et al., 1989). The WHO has a standardized test for washing bednets that is also used for washing clothing (WHOPES, 2005). The method involves cutting four pieces of material and placing them in a shaking water bath with soap, and then subjecting them to two rinses. The material is allowed to dry and then is either washed again or tested for knock-down and mortality by exposing mosquitoes to the washed material. It is important to test all clothing and methods of impregnation with these protocols in order to gain an accurate picture of the protective efficacy and duration of the AI before implementing any intervention.

Conclusions

There is clear evidence that insecticide-treated clothing provides protection against bites from a number of arthropods. Further, there is evidence that insecticide-treated clothing can reduce incidences of disease, but the reduction is likely to depend on the vector targeted and the potential for pathogen transmission. Other factors that should be considered when assessing efficacy against pathogen transmission, as evidenced by the studies reviewed here, include: time of biting; the mechanics behind protection, and the AI utilized. The hours of biting by vectors are important to consider because during the day most people cannot be under the protection of a bednet and may be working in areas such as forests and fields in which vectors are more numerous. Clothing may offer some protection against night-biting mosquitoes, but many factors should be further investigated, such as the timing of wearing the clothing, duration of wear, and residual activity on the skin after the clothing has been removed.

An understanding of the mechanism behind the protection provided by insecticide-treated clothing would help us to determine the most appropriate and efficacious situations in which to use treated clothing as an intervention. For instance, in areas of insecticide resistance the difference between behavioural and physiological protection becomes paramount to the success of an intervention. The true mechanism by which protection is accomplished has yet to be defined. Schreck & McGovern (1989) investigated the combination of clothing with skin repellents. The study demonstrated that mosquitoes will bite areas of the body that are not covered by treated clothing when a skin repellent is not used. When a repellent is used in combination with treated clothing, almost complete protection is achieved. By contrast, when untreated clothing is worn, mosquitoes will bite through the clothing regardless of whether a skin repellent is worn along with the untreated clothing. This suggests that treated clothing is not necessarily ‘repellent’, but it does prevent biting on areas covered by the treated material. Further investigations are needed to confirm whether permethrin-treated clothing can repel mosquitoes and the mechanisms underlying this action. The distinction between repellent and insecticidal properties is an important factor to consider, especially in areas of permethrin resistance. If treated clothing does actively repel arthropods, it may provide protection even in places where physiological resistance to pyrethroids is found. Although not discussed in detail in this review, other repellents, such as those normally applied to the skin (e.g. DEET, PMD and KBR3023), could be used in contexts of permethrin resistance (Table 1).

Under the correct circumstances, insecticide-treated clothing is a promising intervention against disease transmission, but further studies are needed to define the most appropriate guidelines for testing insecticide-treated clothing. Although guidelines will allow for standardized efficacy testing, the use of insecticide-treated clothing should be carefully considered in each individual field setting. At the London School of Hygiene and Tropical Medicine, we are currently testing permethrin-impregnated uniforms under laboratory conditions to measure the efficacy of different types of treatment, including waning efficacy after washing and the behavioural response of permethrin-resistant mosquitoes to treated fabric (Wilder-Smith et al., 2012a). Furthermore, at the University of Bangkok in Thailand, a randomized, controlled, school-based trial is currently underway to establish the efficacy of impregnated uniforms in school-age children in terms of reducing dengue infection (Wilder-Smith et al., 2012b). Cost-effectiveness studies and studies on the longterm safety of impregnated clothes are being performed to assess the value of impregnating clothes or uniforms. For all future studies, monitoring of the effectiveness of the AI over time in real-world settings should provide valuable insights into the practicality of its longterm use in a field setting.

Acknowledgements

We thank Professor Steve Lindsay, School of Biological and Biomedical Sciences, Durham University, Durham, U.K., for his support and advice in the writing of this review. The authors have received funding from the European Union Seventh Framework Programme FP7/2007–2013 under grant agreement no. 282589

Author's declaration of interests

No competing interests have been declared.