Permethrin-treated baby wraps for the prevention of malaria: Uganda

Abstract

Background

Malaria remains a major cause of childhood mortality in sub-Saharan Africa. We sought to leverage the traditional practice of mothers carrying children on their backs in cloth wraps, hypothesizing that treatment with a repellent might provide an additional layer of protection.

Methods

We conducted a double-blind randomized, placebo-controlled trial of permethrin-treated wraps in Uganda. Eligible participants were adult women and their children, age 6 to 18 months. Pairs were randomly assigned (1:1) to receive permethrin-treated or sham-treated wraps. Wraps were retreated every 4 weeks. All participants received a new bednet. Participants attended study clinics every two weeks for six months and made unscheduled visits for febrile illness. The primary outcome was clinical malaria in children, defined as fever and a positive malaria rapid diagnostic test.

Results

Between June 2022 and April 2024, 419 mother-infant pairs were screened and 400 were randomly assigned to permethrin-treated (n=200) or sham-treated (n=200) wraps. Clinic attendance rates were high (5,194 of 5,200, 99.9%), and no participants were lost to follow-up. Clinical malaria incidence was 2.14 (95% CI 1.73–2.62) and 0.73 (95% CI 0.51–1.02) per 100 person weeks in the sham-treated and permethrin-treated arms, respectively (incidence rate ratio 0.34, 95% CI 0.23–0.51, p<0.001). Rash was reported more often in the permethrin-treated (8.5%) than in the sham-treated arm (6.0%).

Conclusions

Use of permethrin-treated baby wraps significantly reduced clinical malaria incidence among young children with access to bednets. Future research should corroborate the efficacy of wraps in different settings and evaluate long-term repellent applications.

(Funded by the Doris Duke Foundation, and others; ClinicalTrials.gov NCT05391230).

Despite gains in recent decades, progress against malaria has stalled. In 2023 there were nearly 600,000 deaths from malaria.¹ Approximately 95% of these deaths occurred in the African region with most occurring among children under five years of age.¹ Multiple factors have been cited, including insecticide resistance² and increases in outdoor biting during dawn and dusk hours.^3,4 Seasonal malaria chemoprevention in which children receive antimalarial medications during peak periods, has proven effective, but its feasibility and sustainability in perennial transmission settings is less certain.^5,6 Malaria vaccines may address these gaps. Yet existing vaccines are characterized by waning protection and appear to be less effective in infants.^7–9 Thus, complementary innovations are needed.

In much of sub-Saharan Africa, mothers carry young children on their backs utilizing wraps made from locally purchased cloth (Figure 1). The wrap, called a lesu in western Uganda, is typically used to carry children from infancy to two years of age. It also serves as a blanket or swaddle when children are set on the ground or put to bed. Thus, both mother and child spend much of the day in contact with the cloth.

Figure 1:

Soaking wraps in 0.5% permethrin (top left); wraps soaking in polyethylene bags (bottom left); wraps in use by study participants (center, right)

We hypothesized that treating wraps with repellent might provide an additional layer of protection against malaria. Similar approaches have been employed for military uniforms^10,11 as well as clothing and blankets for refugees.^12,13 Treatment of wraps has potential advantages including: (i) targeting young children who are most vulnerable to malaria, (ii) integration with existing cultural norms, (iii) offering protection against outdoor and daytime biting mosquitoes, and (iv) minimizing direct contact with permethrin because the wrap is worn over clothing.

METHODS

Study Design

We conducted a double-blind, randomized, placebo-controlled trial in western Uganda. The study area experiences moderate, perennial Plasmodium falciparum transmission with seasonal peaks.^14,15 The primary mosquito vectors are An. gambiae and, to a lesser degree, An. funestus. Across Uganda, the prevalence of genetic markers of pyrethroid resistance exceed 90%.¹⁶ There were no indoor residual spraying or Seasonal malaria chemoprevention campaigns in the study area during the trial period and malaria vaccines were not yet available.

The trial design and protocols have been published¹⁷ and are included in the Supplementary Appendix. The trial was approved by the WCG Institutional Review Board, the Mbarara University of Science and Technology Research Ethics Committee, and the Uganda National Council of Science and Technology. Written informed consent was obtained from the mother. Trial oversight was provided by the Data Safety and Monitoring Board at the North Carolina Translational & Clinical Sciences Institute. All the authors vouch for the accuracy and completeness of the data and for the fidelity of the trial to the protocol (available at NEJM.org). This report adheres to the guidelines for Consolidated Standards of Reporting Trials.¹⁸

Participants

We enrolled women 18 years and older with a child between the ages of six to 18 months. Exclusion criteria were (i) HIV infection or sickle cell disease (because Ugandan guidelines recommend malaria prophylaxis in these populations),^19,20 (ii) known pyrethroid allergy, and (iii) participation in another malaria-related treatment study. Our protocol was amended during the study to additionally exclude mothers with twins due to challenges using the wrap.

Recruitment, randomization and masking

We recruited participants in partnership with local community health workers. After providing written informed consent, mother and child underwent testing for malaria with a histidine rich protein-2 (HRP2) based malaria RDT. Those testing positive received treatment with artemether-lumefantrine in accordance with Ugandan guidelines.²⁰

Participants were randomized to either the permethrin- or sham-treated wrap arms in a 1:1 ratio. Randomization was stratified by age of the child (age 6–11 months and 12–18 months) and clinical site, using a permuted block design. Randomization was performed on-site by a research assistant using REDCap.²¹ Study participants and clinicians remained blinded to the study allocation.

Interventions

At the initial study visit (week 0), participants randomized to the intervention arm received two wraps that had been treated with permethrin, while those in the control arm received sham-treated wraps. We chose permethrin based on its demonstrated efficacy, its commercial availability, and established safety record.²² Swathes of cotton wax cloth measuring 2 × 1 meters, were obtained from a local vendor based on information gained during pilot testing.²³ Wraps were soaked in permethrin diluted to 0.5% in accordance with the manufacturer’s instructions (Sawyer Products, Safety Harbor, FL) and allowed to dry (Figure 1). Participants randomized to the control arm received wraps that underwent a sham treatment by soaking in water. There was no difference in the material or appearance of wraps distributed between arms. However, permethrin does have a distinct odor which could not be blinded, Retreatment occurred in both arms every four weeks during study visits. All participants received a new, pyrethroid-only LLIN (PermaNet 2.0, Vestergaard).

Participants attended study visits every two weeks for 24 weeks. During each visit, participants completed brief health questionnaires. A clinical officer performed an examination of the child, including a dedicated skin examination. If the child was febrile during the visit or had a reported fever in the past 24 hours, they were tested for malaria with a RDT. Children with a positive RDT result were treated in accordance with national guidelines for uncomplicated or severe malaria as determined by the clinical officer.²⁰ Children with a negative RDT were assessed by the clinical officer in accordance with the local standard of care. Participants kept pictorial logs recording the number of nights the child slept under the LLIN. At the scheduled mid-point (week 12) and endpoint (week 24) visits, participants (i) submitted logs documenting the frequency of wrap use and washing from the prior two weeks and (ii) provided capillary blood obtained via finger-prick, which was tested for hemoglobin (Hb) concentration. At the final visit, participants also completed a semi-structured interview eliciting perceptions of and experience with the wraps.

Participants were asked to present to clinic for acute febrile illnesses occurring between study visits. Those who did underwent clinical evaluation as described above. Participants received incentives and transportation stipends for clinic attendance and trial completion.

Outcomes

The primary outcome was clinical malaria in children, defined as fever (temperature of ≥38.0 Celsius) or reported fever within the past 24 hours and a positive malaria RDT. Secondary outcomes were: (1) change in child’s Hb levels; (2) change in children’s growth parameters; (3) change in children’s nutritional status; (4) hospitalization for malaria in children; (5) asymptomatic parasitemia in children; (6) change in the mother’s Hb level; and (7) clinical malaria in the mother. Molecular testing for asymptomatic parasitemia and analysis of maternal outcomes is ongoing and will be reported separately. Safety outcomes included skin reactions to wraps reported by the mother or documented by the clinical officer, hospitalization, or death from any cause.

Statistical analysis

A total of 400 mother-infant pairs (200 per arm) were required to detect a 30% reduction in the incidence of clinical malaria (i.e., incidence rate ratio [IRR]=0.70) in the permethrin-treated arm relative to the sham-treated arm over the 24-week study period with a two-sided test, power of 0.80, and type I error 0.05.²⁴ Power calculations assumed clinical malaria incidence of 3.0 cases per 100 person-weeks in the sham-treated arm.²³ Analysis followed a pre-specified statistical analysis plan¹⁷ with an intention-to-treat approach. A detailed description of the statistical methods is provided in the Supplementary Appendix.

Incidence rates of clinical P. falciparum malaria in children over the 24-week follow-up period were calculated by arm and compared by estimating the IRR and incidence rate difference (IRD) with adjustments for stratification variables. Infections detected during the household visit and baseline clinic visit, prior to wrap distribution, were not included. Time at risk accounted for censoring of the 28-day period following infection^25,26 Exploratory analyses were conducted by age category. Cumulative incidence curves were estimated for each randomization arm, and risk ratios contrasting randomization arms were estimated at 12 and 24 weeks.

Secondary outcomes, including change in child’s Hb level and growth parameters indicative of acute (e.g., mid-upper arm circumference [MUAC]) and chronic (e.g., height-for-weight) malnutrition were evaluated by comparing mean changes between randomization arms from baseline to weeks 12 and 24. The proportion hospitalized in each randomization arm was compared. Sensitivity analyses for primary and secondary outcomes are presented in the Supplementary Appendix. For secondary outcomes, 95% confidence intervals are presented. These intervals have not been adjusted for multiple comparisons and should not be interpreted as hypothesis tests. Adverse reactions were summarized by arm.

RESULTS

Trial Population

Between June 2022 and April 2024, 419 women registered interest, six (1.5%) of whom were ineligible. Study staff conducted 413 household visits, during which 13 (3.3%) women tested positive for HIV and were excluded (Figure 2). The remaining 400 (95.5%) woman-infant pairs were randomly assigned to the permethrin- (n=200) or sham-treated (n=200) arm. All participants received two wraps during the baseline visit (week 0). Two participants, one each in the control and intervention arms, reported losing one of their wraps, which were replaced at the next study visit. Baseline demographic, household, and clinical characteristics are presented in Table 1. Demographic characteristics of participants were similar to those in the general pediatric population of rural East Africa (Supplementary Material). A total of 41 (10.3%) children tested positive for malaria during the baseline household visit and received treatment with artemether-lumefantrine, including 21 (10.7%) in the sham- and 20 (10.1%) in the permethrin-treated arm. At the initial clinic visit (week 0), 62 (15.5%) children were febrile or reported a fever in the past 24 hours, of whom 8 (12.9%) had a positive RDT result.

Figure 2:

Screening, randomization and follow-up

Abbreviations: SCD = sickle cell disease; HIV = human immunodeficiency virus

Table 1:

Demographic and clinical characteristics at baseline of enrolled children by arm. All values shown are n (%) unless otherwise indicated.

	Untreated Wrap (n=200)	Permethrin-treated wrap (n=200)
Demographic characteristics
Female	94 (47.0)	98 (49.0)
Age in months (Median (Q1, Q3))	11.0 (8.0, 14.0)	12.0 (8.0, 15.0)
Age group
6–11 months	103 (51.5)	98 (49.0)
12–18 months	97 (48.5)	102 (51.0)
Maternal age in years (Median (Q1, Q3))	25.0 (21.0, 30.2)	26.5 (22.0, 32.0)
Maternal parity (Median (Q1, Q3))	3.0 (1.0, 4.0)	3.0 (1.0, 5.0)
Maternal education (highest level)
Primary	122 (61.0)	117 (58.5)
O Level	70 (35.0)	66 (33.0)
A Level	1 (0.5)	2 (1.0)
University/Tertiary	5 (2.5)	10 (5.0)
Missing	2 (1.0)	5 (2.5)
Place of birth
Public health centre	134 (67.0)	129 (64.5)
Public hospital	23 (11.5)	28 (14.0)
Private facility	28 (14.0)	34 (17.0)
Home	14 (7.0)	9 (4.5)
Other	1 (0.5)	0 (0.0)
Household characteristics
Household members (Median (Q1, Q3))	6.0 (5.0, 7.0)	6.0 (5.0, 7.0)
Floor material
Earth/sand	118 (59.0)	129 (64.5)
Concrete/cement/bricks/tiles	82 (41.0)	70 (35.0)
Other	0 (0.0)	1 (0.5)
Primary drinking water source
Piped	74 (37.0)	63 (31.5)
Public tap	43 (21.5)	40 (20.0)
Well or borehole	41 (20.5)	47 (23.5)
Surface water	42 (21.0)	50 (25.0)
Own livestock	91 (45.5)	88 (44.0)
Missing	2 (1.0)	0 (0.0)
Electricity in home	44 (22.0)	32 (16.0)
Missing	1 (0.5)	1 (0.5)
Malaria & Care seeking
Number of LLINs in household (Median Q1, Q3)	1.0 (1.0, 2.0)	1.0 (1.0, 2.0)
Self-reported travel time to nearest health facility in minutes (Median [Q1, Q3])	24.00 (18.0, 28.5)	24.00 (19.5, 28.0)
Sought care in last 2 weeks for fever in child	61 (30.5)	59 (29.5)
Any drug taken in last 2 weeks for fever in child	71 (35.5)	66 (33.0)
Clinical characteristics
MUAC in cm (Median (Q1, Q3))	14.3 (13.8, 15.1)	14.3 (13.6, 15.2)
MUAC categorya
Severe acute malnutrition	0 (0.0)	0 (0.0)
Moderate acute malnutrition	30 (15.0)	37 (19.0)
Adequately nourished	170 (85.0)	163 (82.0)
Hemoglobin in g/dL (Median (Q1, Q3))	11.9 (10.6, 12.9)	12.0 (10.7, 13.0)
mRDT positive at home visit	21 (10.5)	20 (10.0)
mRDT performed at initial clinic visit	35 (17.5)	27 (13.5)
mRDT positiveb at initial clinic visit	5 (14.3)	3 (11.1)

^aSevere acute malnutrition defined as MUAC < 11 cm, moderate acute malnutrition defined as MUAC < 13.5 cm, adequately nourished defined as MUAC >=13.5 cm.

^bDenominator among those tested

Participants completed 5,194 of 5,200 (99.9%) scheduled study visits. No participants were lost to follow-up. Children were reported to have slept under the LLIN more than 99% of nights. Daily wrap use was high throughout the study period and washing occurred at similar frequencies in the untreated and treated wrap arms (Supplementary Material).

Primary Outcome

We performed 741 RDTs for symptomatic illness in participating children. Of these, 128 (17.3%) RDTs were positive for malaria, with 34 (26.6%) and 94 (73.4%) and occurring in the permethrin- and sham-treated arms, respectively. The incidence of clinical malaria was 0.73 and 2.14 cases per 100 person weeks in the permethrin- and sham-treated arms, respectively (IRR 0.34, 95%CI 0.23–0.51, IRD −1.41, 95%CI −2.02 - −0.83, Table 2). Risk of first clinical malaria episode was lower in the permethrin-treated than the sham-treated arm throughout the 24-week follow-up period, with estimated risks at 24 weeks of 0.16 and 0.34, respectively, and a risk ratio of 0.48 (95% CI 0.33 – 0.69) (Figure 3). A total of 21 participants experienced multiple clinical malaria episodes during the trial, including 19 (90.5%) in the sham- and 2 (9.5%) in the permethrin treated-wrap arms.

Table 2:

Incidence of clinical malaria among children stratified by arm

	Sham-treated wrap (n=200)	Permethrin-treated wrap (n=200)	IRDa (95% CI)	IRRa (95% CI)
Clinical malaria episodes	94 (73.4%)	34 (26.6%)	-	-
Time at risk (person-weeks)	4,395	4,641	-	-
Clinical malaria incidence (per 100 person-weeks)	2.14 (1.73, 2.62)	0.73 (0.51, 1.02)	−1.41 (−2.03, −0.83)	0.34 (0.23, 0.51)b
Age 6–11 months	2.07 (1.52, 2.76)	0.66 (0.37, 1.09)	−1.42 (−2.22, −0.70)	0.32 (0.18, 0.57)
Age 12–18 months	2.21 (1.62, 2.94)	0.80 (0.48, 1.26)	−1.41 (−2.39, −0.51)	0.36 (0.21, 0.62)

^aIncidence rate differences (IRD) and incidence rate ratios (IRR) are adjusted for randomization strata (study site and child’s age at the household visit) for overall estimates and study site for age group estimates.

^bp-value <0.001

Figure 3:

Cumulative incidence of first clinical malaria among children

Secondary Outcomes

In total 30 children were admitted to the inpatient ward for treatment of malaria including 10 (5.0%) in the permethrin-treated and 20 (10.0%) in the sham-treated arms (Table 3). Change in Hb levels and markers of acute malnutrition, assessed by MUAC, were comparable between arms at 12- or 24-week visits. Calculated Z-scores for growth measures were similar (Table 3), although the direction across measures was contrary to the hypothesized effect.

Table 3:

Secondary outcome measures

	Sham-treated wrap (n=200)	Permethrin-treated wrap (n=200)	Difference (95% CI)
Participants hospitalized	20, 10.0% (6.2%, 15.0%)	10, 5.0% (2.4%, 9.0%)	−5.0% (−10.7%, 0.8%)
Change in Hemoglobin (g/dL)
Baseline to mid-point (week 12)	0.81	0.76	−0.05 (−0.34, 0.25)
Baseline to study end (week 24)	1.34	1.19	−0.15 (−0.51, 0.22)
Change in MUAC (cm)
Baseline to mid-point (week 12)	0.60	0.65	0.05 (−0.10, 0.20)
Baseline to study end (week 24)	1.27	1.27	0.00 (−0.17, 0.16)
Change in height-for-weight (Z-score)
Baseline to mid-point (week 12)	0.49	0.46	−0.03 (−0.26, 0.20)
Baseline to study end (week 24)	0.78	0.74	−0.04 (−0.31, 0.23)
Change in height-for-age (Z-score)
Baseline to mid-point (week 12)	0.03	−0.02	−0.06 (−0.24, 0.12)
Baseline to study end (week 24)	0.18	0.09	−0.09 (−0.34, 0.16)
Change in weight-for-age (Z-score)
Baseline to mid-point (week 12)	0.45	0.38	−0.07 (−0.22, 0.09)
Baseline to study end (week 24)	0.75	0.65	−0.10 (−0.29, 0.08)

Safety Outcomes

A total of 34 participants experienced AEs including 20 (10.0%) children in the permethrin-treated and 14 (7.0%) children in the sham-treated wrap arms (Supplementary Material). Rash was the most frequently reported event occurring in 17 (8.5%) children in the permethrin-treated and 12 (6.0%) children in the sham-treated wrap arm. All rashes were assessed as mild and none led to study discontinuation. One case of severe malaria was observed in the sham-treatment arm (0.5%, 95% 1-sided upper CI = 2.4%). No SAEs were observed in the permethrin-treated wrap arm.

DISCUSSION

Use of permethrin-treated wraps reduced clinical malaria incidence among young children compared to use of sham-treated wraps in Uganda. This level of protection was attained even with provision of LLINs to participants and high levels of reported use. Adverse events were infrequent, mild, and similar between arms. These findings demonstrate that permethrin-treated wraps are effective at preventing malaria in perennial-transmission settings and without evident safety concerns, at least for six months of use. Given alignment with existing cultural practices, treated wraps could provide an additional level of protection against malaria. Assuming the wraps primarily exerted their effect on dusk/dawn biting, they could work synergistically with household-based interventions such as LLINs and indoor residual spraying. Treated wraps may also complement malaria vaccines, which appear to have reduced efficacy in infants, by bridging children until vaccine-induced immunity is obtained.^7,8

In the trial planning process, we did not have rigorous data informing estimates of the number of infectious bites that occur (i) outside the home or (ii) when the child is inside the home, but not under a LLIN. Yet, the magnitude of the effect found in our study, especially given the reported LLIN use among participants, suggests that a substantial proportion of malaria risk is occurring during these times. These findings highlight the gaps in LLIN-focused strategies and reiterate the need for additional studies of outdoor biting and novel approaches for prevention and control.

Our trial did not incorporate entomological endpoints and therefore cannot confirm the mechanism of effect. We hypothesized that deterring mosquito landing and biting, rather than the killing of mosquitoes, would be the primary means of prevention. Even as Anopheles become resistant to the killing effect of permethrin, they may still be repelled. A recent trial of a pyrethroid-based spatial repellent also found substantial reductions in malaria incidence despite high levels of resistance.²⁷ The repellent effect may also provide some protection to those in the immediate area, including the mother and other household members.

We used a strategy of frequent permethrin re-treatment of the wraps to avoid washout. However, monthly, facility-based re-treatment may be resource-intensive and challenge feasibility at scale. With evidence of efficacy with frequent re-treatment, implementation studies to determine optimal repellent agents, treatment approaches, and frequency of re-treatment are now needed. For example, uniforms impregnated with permethrin at the factory, like those utilized by many militaries, have shown protection against vector-borne disease for up to one year without retreatment.²⁸ Factory-treatment would also have the advantage of tighter bonding to fabric and less absorption through the skin. For example, a review of factory-treated clothing by the US Environmental Protection Agency, including clothing for children and toddlers, did not identify any scenarios of concern.²⁹ Given the anticipated duration and frequency of use, however, extended follow-up of children, particularly for neurodevelopmental impacts of permethrin exposure is needed. Yet malaria infection, both severe and uncomplicated forms, can cause long-term cognitive impairment, and a careful weighing of potential risks versus benefits will be required.³⁰

This trial has several strengths. It was conducted at rural health centers that are representative of facilities where most residents of sub-Saharan Africa receive malaria care. Additionally, by partnering with community health workers, we were able to achieve high rates of participation. There are also important limitations. First, the study is most generalizable to contexts where baby wraps are used to carry young children. Second, our trial took place at one geographic location. Additional research is needed to determine the efficacy of wraps in other contexts including with different malaria vectors and transmission intensities. Third, estimates of LLIN use and frequency of wrap washing were self-reported and may have been subject to social desirability or recall biases. However, there is no reason to believe that such bias would be different between the arms. Although we made efforts to shield participants from knowledge of their assignment, including sham re-treatments, permethrin does have a distinct odor, which could have compromised blinding. This concern is mitigated by surveys of participants that showed high confidence in benefit from wrap usage in both arms (Supplementary Material). Lastly, we did not assess changes in feeding blood-meal seeking behaviors in response to permethrin-treated wraps, although such effects of repellents have not been reported elsewhere.²⁷

In summary, in this randomized, double-blind, placebo-controlled trial, use of permethrin-treated wraps led to substantial reductions in malaria incidence without safety concerns among young children in rural western Uganda. The reduction was seen in the setting of high LLIN use. Additional research is needed to determine optimal permethrin application methods (e.g., factory impregnation) and frequencies to inform implementation strategies and cost-effectiveness analyses. Future studies should incorporate entomological measures to better define the mechanism of effect and longer-term follow up of children to assess for potential neurodevelopmental effects of permethrin exposure.

Availability of Data and Materials

Deidentified individual data that supports the results will be shared beginning 9 to 36 months following publication provided the investigator who proposes to use the data has approval from an Institutional Review Board (IRB), Independent Ethics Committee (IEC), or Research Ethics Board (REB), as applicable, and executes a data use/sharing agreement with UNC.