Community reservoirs of malaria parasites and gametocytes in Arba Minch district, southern Rift Valley, Ethiopia: a cross-sectional study

Abstract

Background

This study aimed to assess the community-based prevalence of malaria reservoirs following cases visiting health facilities. The diagnostic performance of microscopy in detecting community-based malaria parasites was compared to nested polymerase chain reactions (PCR).

Methods

From July to October 2022, reactive case detection was conducted in Sile village, Gamo Zone, in the Southern Rift Valley of Ethiopia. Within six days of identifying an index case, all individuals in the same household and neighboring households were screened for malaria by microscopy, with nested PCR for confirmation. Asexual parasite and gametocyte density were measured microscopically.

Results

Of the 2434 individuals visited following 142 PCR-confirmed index cases, 2009 were included in the final analysis. The PCR-corrected, microscopy-based malaria prevalence in the study community was 3.6% (72/2009; 95% CI 2.8–4.5). Subsequent PCR analysis of randomly selected microscopy-negative samples identified an additional 33 submicroscopic infections, yielding a submicroscopic prevalence of 10.1% (33/326; 95% CI 7.2–13.9). Submicroscopic prevalence was 4.6% for P. vivax (15/326; 95% CI 2.6–7.5) and 4.3% (14/326; 95% CI 2.4–6.9) for P. falciparum. Mixed infections comprised 1.3% (4/326; 95% CI 0.3–3.1) of the cases. Overall, submicroscopic infections accounted for 31% (33/105; 95% CI 22.6–40.8) of the total PCR-confirmed malaria cases in the community, indicating that nearly one-third were missed by microscopic examination. Index cases had higher asexual parasite density (16,177 vs. 1900/μL; P < 0.001) but lower gametocyte carriage than reactive cases, despite similar gametocyte densities (600 vs. 482/μL; P = 0.08). The gametocyte carriage rate was higher among P. vivax (22/32; 69%) than among P. falciparum (6/27; 22%) reactive cases.

Conclusion

The high gametocyte carriage rate among microscopy–reactive cases highlights the potential role of community-based infections in sustaining malaria transmission.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12936-026-05795-2.

Background

Up to 2017, malaria control in Ethiopia improved. Insecticide-treated bed nets, indoor residual spraying, and prompt diagnosis and treatment of malaria cases were the main strategies. However, since 2023, there has been a surge in malaria cases in the country [1], and the number of cases has surpassed 12 million in 2024 [2].

In regions where malaria is prevalent, the rise of submicroscopic and asymptomatic cases creates obstacles to controlling the disease [3]. Individuals without symptoms harbor the parasite, remain infectious, and can contribute to the under-recognized malaria burden, mainly because of gametocytes in asymptomatic and thus untreated cases [4]. However, the magnitude of this problem has not been well studied in Ethiopia.

This study highlights the burden of malaria cases and the potential for transmission beyond those detected at health facilities, emphasizing the role of subclinical infections within communities. In areas with a high prevalence of gametocyte carriers, effective control strategies must go beyond standard symptomatic treatment to directly target these hidden community reservoirs [5]. Gametocyte levels differ between symptomatic and asymptomatic malaria cases, shaping transmission dynamics [6, 7]. Since gametocytes are the source of infection for mosquitoes, it is vital to assess their prevalence both in index cases (patients at health facilities) and in reactive cases (within the surrounding community). Measuring gametocyte carriage in reactive cases helps estimate local transmission potential. Additionally, because gametocytes arise from schizonts, the asexual parasite stage must also be considered in understanding transmission [8].

One method for assessing remaining malaria cases in the community is to follow cases visiting health facilities, screening individuals in their households and neighbouring areas [9]. This strategy helps identify reservoirs and provide treatment to prevent further spread of the disease by breaking transmission chains. We hypothesize that by conducting such surveillance, malaria programs can quickly stop outbreaks, reduce parasite reservoirs, and protect communities, ultimately lowering the overall burden of malaria. However, such an approach requires diagnostic accuracy, as an inaccurate diagnosis can leave some cases in the community that may lead to transmission [10]. Enhanced diagnostic capacity facilitates timely treatment, reduces malaria transmission, and strengthens efforts toward malaria control and eventual elimination. Hence, this study aims to assess the prevalence of malaria reservoirs within the community, focusing on parasite and gametocyte densities and carriage rates to better understand their contribution to persistent malaria transmission.

Materials and methods

Study area description

The study was conducted in Ganta Kanchama Kebele, specifically in Sille village, Arba Minch area, Gamo Zone, southern Ethiopia (Fig. 1). The village was selected for its high malaria endemicity and proximity to Arba Minch town, which enabled immediate malaria microscopy following field screening. The village is 518 km southwest of Addis Ababa, the capital of Ethiopia, and 13 km from Arba Minch, the capital of the Gamo Zone. The altitude ranges from 1120 to 1380 m above sea level. It is situated at 5°99′N latitude and 37°50′E longitude. Its mean annual rainfall ranges from 900 to 1300 mm, and its temperature ranges from 25 to 36 °C. The village is adjacent to Lake Chamo, one of Ethiopia's southern Rift Valley lakes.

Fig. 1.

Map of the study area

(Source: CSA-2007)

There is a health post in the village that provides primary healthcare, including malaria diagnosis using a rapid diagnostic test (RDT), and provides free antimalarial medications. The residents' livelihoods depend on farming and fishing. The primary economic source is planting maize, mango, and bananas, mainly by irrigation. Additionally, growing tomatoes and papayas is becoming a common practice in the village. The water sources for irrigation are the Sille River and Lake Chamo.

Malaria is endemic in the village. Anopheles arabiensis and An. pharoensis are malaria vectors in the village [11].

Study design and sampling

This cross-sectional study was conducted from July to October 2022. The original plan was to use Rapid Diagnostic Tests (RDT) to diagnose malaria. Unfortunately, after we initiated the study, the health post did not receive RDTs, and the initial diagnosis was based solely on microscopy. Index cases were first detected at the health post. Field teams then visited the index household and neighbouring households to collect blood smears and demographic data.

The screening in the community was done by microscopic examination of Giemsa-stained blood smear slides within six days of case identification at the health post. Slides were processed and examined within 24 h to ensure timely diagnosis and treatment. Individuals who tested positive were treated according to the Ethiopian Federal Ministry of Health’s malaria treatment guidelines [12]. Sampling stopped once reactive case clusters began to overlap, and index cases were confirmed to originate throughout the study site. All positive samples, along with randomly selected samples that tested negative by microscopy, were confirmed using nested PCR.

Eligibility criteria

The study involved 2435 household members from the 142 index cases. Only permanent residents of the village were included in the survey; daily laborers who commuted to home each day were not considered.

Blood sample collection and microscope examination

Thick and thin blood films were prepared on a single pre-labelled, frosted-end slide to detect, identify, and quantify Plasmodium parasites. The thin smears were fixed in absolute methanol for 10 to 20 s, and both thick and thin films were stained with a freshly prepared 10% Giemsa solution for 10 min. After air-drying, the slides were examined under oil immersion with a 100 × objective lens. A slide was recorded as negative if no Plasmodium parasites were observed in 200 microscopic fields.

Parasite density was quantified against 200 white blood cells (WBC) if ≥ 100 parasites were counted or against 500 WBCs if ≤ 99 parasites were counted. Counting additional WBCs improves the accuracy of parasitemia detection at low parasite densities. Gametocyte density was measured by examining Giemsa-stained thick blood smears and counting gametocytes per 200 or 500 WBCs, depending on parasite density. A standard WBC count of 8000 leukocytes/μl was used to quantify sexual and asexual parasites [13]. Both asexual and sexual stages of Plasmodium parasites were counted separately. An experienced malaria microscopist first examined each slide. A second, blinded microscopist independently reviewed the same slide. In cases of discordant results, a third expert microscopist conducted a final review to resolve discrepancies.

Simultaneously with the collection of blood smears, dried blood spots (DBS) were prepared from a finger prick, with three spots applied to each card to fill the designated circles. After air-drying for 3 h on a flat surface, the cards were placed in sealable plastic bags with a desiccant and transported to the Advanced Medical Entomology and Vector Control Laboratory at Arba Minch University for storage at − 40 °C for PCR analysis.

DNA extraction and detection of Plasmodium parasites using nested PCR

Genomic DNA was extracted from DBS samples collected from study participants using the Chelex-saponin (Sigma-Aldrich, Merck KGaA, Darmstadt, Germany) method [14]. One-punch DBS sample (diameter 6 mm and ≈20 µL of whole blood) from the patient was digested with a 0.5% Chelex-saponin buffer, and nested PCR was performed to detect and identify Plasmodium species using standardized protocols [15]. The first round of DNA amplification was performed using rPLU5 (CCTGTTGTTGCCTTAAACTTC) and rPLU6 (TTAAAATTGTTGCAGTTAAAACG) primers to determine the Plasmodium genus. N1 PCR reactions were performed according to manufacturers’ guideline: 5 µl of DNA extract was combined with 5 × PCR buffer (Promega Corporation, Madison, WI 53711 USA), 25 mM magnesium chloride, 10 μM deoxynucleoside triphosphates (dNTPs), 10 μM forward and reverse primers (rPLU5 CCTGTTGTTGCCTTAAACTTC and rPLU6 TTAAAATTGTTGCAGTTAAAACG), 5 unit Taq DNA polymerase (Thermo Fisher Scientific Inc.) and 25 µL nuclease-free water. Thermal cycling was performed on a Bio-Rad PCR System (Bio-Rad, Hercules, California, United States): 10 min initial denaturation at 95 °C, followed by 35 cycles at 95 °C for 10 s, 58 °C for 60 s, and 72 °C for 90 s, with a final 10 min extension step at 72 °C.

All positive samples from N1 subjected to the N2 to identify species-specific genes: P. vivax (120 bp) and P. falciparum (205 bp). 2 µL aliquot from the product of the N1 was used as template for Plasmodium species-specific fragment amplification using two pairs of primers (rFAL-F-TTAAACTGGTTTGGGAAAACCAAATATATT) and (rFAL-R-ACACAATAGACTCAATCATGACTACCCGTC) for P. falciparum and (rVIV1 (5′CGCTTCTAGCTTAATCCACATAACTGATAC3′) and rVIV2 (5′ACTTCCAAGC-CGAAGCAAAGAAACTCCTTA3′) for P. vivax. The thermal cycling conditions were identical, with the number of cycles reduced to 30.

All products were separated by 2% agarose (Thermo Fisher Scientific Inc.) gel electrophoresis in 1X Tris-Borate-Ethylenediaminetetraacetic acid (EDTA) (Thermo Fisher Scientific Inc.) at 120 V for 45 min. DNA was stained in-gel with 8.2 µl GelRed (Merck KGaA, Darmstadt, Germany) and visualised under ultraviolet illumination.

PCR-confirmed positive samples were used as positive controls. For negative controls, an empty DBS card was used for extraction, and PCR-grade water was used for the master mix.

Data analysis

Statistical analysis was performed using R Studio version 2024.04.2. Cross-tabulations and bar plots were used to illustrate the distribution of test results across key demographic and clinical variables. Gametocyte carriage rates were analyzed using proportion tests and visualized through annotated bar plots. Violin plots, based on log-transformed data, were employed to compare asexual and gametocyte densities across case categories and species. Group differences were assessed using Wilcoxon rank-sum tests, and effect sizes were quantified using Cliff’s Delta to interpret the magnitude of observed differences. Diagnostic performance was evaluated using PCR as the gold standard. A multiclass evaluation was conducted to assess species-specific diagnostic accuracy, and Kappa statistics to demonstrate agreement between microscopy and PCR results. Statistical significance was declared at a p-value threshold of ≤ 0.05.

Results

Socio-demographic characteristics of the study participants

A total of 184 suspected malaria index cases were identified, with 142 confirmed by PCR. Among 2434 nearby residents identified, 425 were excluded after three visits. Blood samples were collected from 2009 individuals, resulting in 82 microscopy-positive and 72 PCR-confirmed cases.

Of the 1927 microscopy-negative individuals, 326 were randomly selected for PCR testing, of whom 33 tested positive. The flowchart is presented in Fig. 2.

Fig. 2.

Study flow chart

Of the 2009 participants, children under 5 years accounted for 17.6% (354/2009). Children aged 5–14 years represented the largest proportion, making up 43.3% (n = 870). Individuals older than 14 years constituted 39.1% (n = 785). Female respondents made up a higher proportion, at 59.5% (n = 1196). In total, 3% (n = 61) of the participants reported experiencing a fever within the 48 h preceding the survey, whereas only 0.4% (n = 9) presented with a measured fever (axillary temperature ≥ 37.5 °C) at the time of the survey. The majority (97.0%, n = 1948) reported no signs or symptoms. A small percentage, 4.2% (n = 84), reported sleeping under a bed net (Table 1).

Table 1.

Socio-demographic characteristics of reactive cases in Sille village, South Ethiopia (July to October 2022)

Characteristics	Category	N (%)
Age	< 5	354 (17.6)
	5–14	870 (43.3)
	> 14	785 (39.1)
Sex	Male	813 (40.5)
	Female	1196 (59.5)
Sign and symptom	Yes	61 (3.0)
	No	1948 (97.0)
Sleep under bed net	Yes	84 (4.2)
	No	1925 (95.8)
Total study individual	–	2009

Reactive malaria prevalence

The PCR-corrected, microscopy-based malaria prevalence in the study community was 3.6% (72/2009; 95% CI 2.8–4.5). PCR analysis of microscopy-negative samples revealed a submicroscopic prevalence of 10.1% (33/326; 95% CI 7.2–13.9). Submicroscopic prevalence was 4.6% for P. vivax (15/326; 95% CI 2.6–7.5) and 4.3% (14/326; 95% CI 2.4–6.9) for P. falciparum. Mixed infections of the P. falciparum and P. vivax comprised 1.3% (4/326; 95% CI 0.3–3.1) of the cases. Among the malaria cases identified, 31% (33/105; 95% CI 14–29.8) were submicroscopic infections missed by microscopy but detected by PCR testing.

Of the 105 malaria-positive cases identified, 44.8% (47/105; 95% CI 35.5–54.4) were P. vivax, 39% (41/105; 95% CI 30.0–48.8) were P. falciparum, and 16% (17/105; 95% CI 10.2–24.3) were mixed infections (Table 2).

Table 2.

The PCR-corrected estimates of malaria prevalence in a village near Lake Chamo, South Ethiopia (July to October 2022)

Category	No. diagnosed	Pf	Pf % (95% CI)	Pv	Pv % (95% CI)	Mixed	Mixed % (95% CI)	Total cases	Total % (95% CI)
< 5 years	354	12	3.4 (1.8–6.2)	9	2.5 (1.2–4.8)	6	1.7 (0.6–3.6)	27	7.6 (5.1–10.9)
5–14 years	870	19	2.1 (1.3–3.3)	25	2.9% (1.9–4.3)	6	0.7 (0.3–1.5)	50	5.7 (4.3–7.5)
> 14 years	785	10	1.3 (0.6–2.4)	13	1.7 (0.9–2.9)	5	0.6 (0.2–1.5)	28	3.6 (2.4–5.1)
Male	1196	17	1.4 (0.8–2.3)	23	1.9 (1.2–2.9)	6	0.5 (0.2–1.1)	46	3.8 (2.8–5.1)
Female	813	24	3.0 (1.9–4.4)	24	3.0 (1.9–4.4)	11	1.4 (0.7–2.5)	59	7.3 (5.6–9.3)
Symptomatic	61	10	16.4 (7.1–25.7)	7	11.5 (4.7–22.2)	5	8.2 (2.7–18.1)	22	36.0 (23.6–49.8)
Asymptomatic	1948	31	1.6 (1.0–2.1)	40	2.0 (1.4–2.7)	12	0.6 (0.3–1.1)	83	4.3 (3.4–5.3)
Use a bed net	84	2	2.4 (0.0–5.7)	2	2.4 (0.0–5.7)	1	1.2 (0.0–6.5)	5	6.0 (1.9–13.5)
No bed net	1925	39	2.0 (1.4–2.6)	45	2.3 (1.7–3.1)	16	0.8 (0.5–1.3)	100	5.2 (4.2–6.3)

Parasite density of malaria cases within the community

Malaria index cases had higher median parasite densities (16,177 parasites/μL) compared to reactive cases (1900 parasites/μL); (W = 6426, δ = 0.587, P < 0.00001) (Fig. 3).

Fig. 3.

Violin plots illustrate the density of asexual parasite densities for P. falciparum and P. vivax, as detected by microscopy. The width and shape of each violin reflect the frequency and variability of parasite densities within each group

Index cases of P. falciparum had a significantly higher median parasite density of 20,320 parasites/μL compared to reactive cases (4400 parasites/μL) (P < 0.001). A similar trend was observed for P. vivax infections, with median densities of 7900 parasites/μL in index cases, compared to 1428 parasites/μL in reactive cases (P < 0.001) (Table 3).

Table 3.

Comparison of median parasite and gametocyte densities between index cases and reactive cases in a village close to Lake Chamo, South Ethiopia (July to October 2022)

Category	Variable	Median (parasite/μL)	IQR (Parasite)	P-value	Median (gametocyte/μ)	IQR (Gametocyte)	P-value
Overall	Index vs reactive	16,177 vs. 1900	4810–35,700 vs. 469–8050	< 0.001	600 vs. 482	232–1553 vs. 115–1600	< 0.08
Age < 5 years	Index	10,500	1550–23,250	< 0.05	800	400–2514	< 0.05
	Reactive	2110	910–4174		320	160–1230
Age 5–14 years	Index	14,320	4393–21,500	< 0.05	3334	783–5613	< 0.05
	Reactive	1720	357–6796		1600	150–1900
Age > 14 years	Index	18,300	6332–42,810	< 0.05	360	176–1136	< 0.05
	Reactive	1800	288–8430		100	80–231
Male	Index	16,440	5680–31,720	< 0.05	520	240–1004	0.12
	Reactive	1640	478–8832		700	210–2050
Female	Index	15,200	4686–44,220	< 0.05	800	232–2122	0.15
	Reactive	2133	480–6222		362	85–1507
P. falciparum	Index	20,320	7760–45,705	< 0.001	376	188–790	< 0.05
	Reactive	4400	704–8,845		1815	458–3150
P. vivax	Index	7900	1584–15,600	< 0.001	1032	330–3440	< 0.05
	Reactive	1428	237–3775		360	105–1600

Gametocyte density of reactive malaria cases

Overall, the gametocyte densities observed in index cases (600 gametocytes/μL) were comparable to those in reactive cases (482 gametocytes/μL; P = 0.08). However, species-specific analysis revealed that index cases of P. falciparum had significantly lower gametocyte densities (376 gametocytes/μL) compared to the reactive cases (1815 gametocytes/μL; P < 0.05). Conversely, P. vivax infections exhibited the opposite trend, with index cases showing higher gametocyte densities than reactive cases (P < 0.05) (Fig. 4; Table 3).

Fig. 4.

Violin plots illustrate the distribution of gametocyte densities for P. falciparum and P. vivax, as detected by microscopy. The width and shape of each violin reflect the frequency and variability of gametocyte densities within each group

Gametocyte carriage rates among index and reactive cases

Reactive cases exhibited a significantly higher overall rate of gametocyte carriage compared to index cases, with rates of 47.5% (28/59) versus 25.3% (35/138), respectively (P < 0.0039). However, there was no statistically significant difference (P = 0.4664) in the gametocyte carriage rates between reactive cases (69%; 22/32) and index cases of P. vivax (57%; 20/35). Likewise, although the gametocyte carriage rate for P. falciparum was higher in reactive cases (22%; 6/27) compared to index cases (15%; 15/103), this difference was not statistically significant (P = 0.5036) (Fig. 5).

Fig. 5.

The gametocyte carriage rates of index (health facility) and reactive (community) cases in Sile village, Gamo zone, south Ethiopia

Discussion

The study revealed that the submicroscopic infection was high in the in the community and nearly one-third of the malaria cases in the community were missed by microscopic examination. The study also found that gametocyte carriage was higher among community cases (reactive) than among those from health facilities (index). Notably, gametocyte densities were higher in community-identified P. falciparum cases compared to index cases, whereas the reverse was observed for P. vivax, with index cases exhibiting higher gametocyte densities.

Our findings indicate that a substantial proportion of malaria cases were submicroscopic, underscoring the need for more sensitive molecular tools, such as PCR. PCR can detect parasitemia as low as 0.5 parasites/μL, making it a critical diagnostic tool for such cases [16]. In Ethiopia, reactive case detection (RCD) consistently identifies hidden reservoirs of low-density infections [17, 18]. Similarly, in Príncipe (West Africa), qPCR identified over 90% of low-density infections missed by conventional diagnostics during RCD [19]. Despite their low parasite density, these infections may persist untreated and contribute to ongoing transmission [20].

The gametocyte carriage rate was higher among reactive cases compared to index cases. This finding aligns with studies from Kenya [21], which also reported elevated gametocyte carriage rates among asymptomatic community cases. Similarly, a study in northern Ethiopia documented a gametocyte prevalence of 43.3% [22], closely mirroring the rate observed in the current study. In endemic regions, repeated malaria exposure fosters immune tolerance, suppressing clinical symptoms while allowing low-level parasitemia and gametocytemia. For instance, the Ugandan study found that individuals with partial immunity had a 2.7-fold higher rate of gametocyte detection [6]. Conversely, in low-transmission settings such as Thailand, residual infections tend to be chronic and gametocyte-rich [23]. Similarly, research in Burkina Faso demonstrated that chronic asymptomatic infections exhibited significantly higher gametocyte densities and mosquito infectivity than symptomatic infections [24].

Our study observed a higher prevalence of gametocyte carriage in P. vivax infections compared to P. falciparum cases detected through reactive surveillance. The variation likely reflects intrinsic biological differences between the two species. In P. vivax, gametocytes appear early in infection, often before symptoms develop [25]. In contrast, P. falciparum gametocytes mature lately and remain in the bone marrow before entering the bloodstream [6]. This sequestration may explain why index cases (early-detected infections) exhibit lower gametocyte densities compared to reactive cases, which could represent later-stage infections with mature gametocytes. Furthermore, P. falciparum reactive cases may reflect older infections with higher cumulative gametocyte production due to prolonged parasite survival, consistent with evidence that longer-lasting infections are more likely to generate gametocytes [26, 27]. Another contributing factor is P. vivax’s unique ability to form dormant liver-stage hypnozoites, enabling recurrent infections—a mechanism absent in P. falciparum [28]. Despite generally lower parasite densities, P. vivax maintains a substantial gametocyte reservoir, sustaining transmission even as P. falciparum prevalence declines [27].

The 10.1% positive cases were missed by microscopy, underscoring a critical limitation of microscopy in community-based surveillance, a challenge well documented across multiple epidemiological settings [29]. This diagnostic gap is particularly pronounced in low-transmission regions, where parasite densities are frequently below the detection threshold of conventional light microscopy. For instance, a recent study conducted in Ethiopia’s low-transmission areas revealed that microscopy identified only 6.3% of asymptomatic infections, whereas nested PCR (nPCR) detected 20.3% [30]. In asymptomatic individuals with chronic, low-level parasitemia act as silent reservoirs, sustaining transmission even in areas nearing elimination. This phenomenon complicates malaria control efforts, as standard diagnostic tools fail to identify a substantial proportion of infections that contribute to ongoing transmission. Strengthening surveillance among asymptomatic individuals through active case detection and confirmatory testing is crucial for progressing toward malaria elimination goals.

This study has several limitations that should be taken into account when interpreting the findings. First, the single-season and single-site study limits the generalizability of the results, as malaria transmission can vary by season and geographic location. The subgroup analyses—particularly for mixed-species infections and certain age groups—were limited by small sample sizes. While microscopic negative samples were randomly selected for PCR analysis, many remained untested, potentially leading to an underestimation of submicroscopic malaria prevalence. The gender imbalance within the sample (i.e., the predominance of male participants) constitutes a further limitation of the study.Despite these limitations, the study offered valuable insights into the diagnostic challenges and hidden malaria reservoirs within the community, highlighting the need for more sensitive and accurate diagnostic tools for screening malaria cases.

Conclusions

Reactive cases infections represent important reservoirs of malaria transmission due to high gametocyte carriage despite lower parasite densities. Integrating molecular diagnostics into community-based case detection enhances identification of asymptomatic and low-density infections, thereby strengthening surveillance, guiding targeted treatment, and enhancing malaria control in endemic settings such as Ethiopia.

Abbreviations

PCR

Polymerase chain reaction

RDT

Rapid diagnostic test

CSA

Central statistical agency

WBCs

White blood cells

DBS

Dried blood spot

RCD

Reactive case detection

qPCR

Quantitative polymerase chain reaction

nPCR

Nested polymerase chain reaction

Author contributions

ZZ and FM Conceived and designed the study, ZZ, YT, and FM Analysed and interpreted the data, ZZ and NE Conducted laboratory analysis, FM drafted the paper, ZZ, NE, YT, and BL Revised the manuscript, approved the final version. FM and BL Supervised the field and laboratory work. All authors read and approved the final manuscript.

Funding

This study received funding from the Norwegian Programme for Capacity Development in Higher Education and Research for Development (Grant No. QZA-21/0162).

Data availability

The datasets analysed in this study are included in the manuscript and provided as supplementary files.

Declarations

Ethics approval and consent to participate

Ethical approval was obtained from the Arba Minch University IRB (IRB/1292/2022). Written informed consent was obtained from all adult participants and from the parents or guardians of children under 12. Children under 18 provided oral assent. Participation was voluntary and confidential. All individuals who tested positive for malaria were treated according to national guidelines at the local health post.

Competing interests

The authors declare no competing interests.