Epidemiological characteristics, prevention and control status, and factors associated with malaria infection among overseas Chinese workforce in Uganda

Abstract

Background

Malaria remains a major global public health challenge, particularly in sub-Saharan Africa. Overseas Chinese workforce in Uganda face an increasing risk of malaria infection. This study aimed to characterize the epidemiological features and current status of malaria prevention and control among this population, and to identify factors associated with malaria infection.

Methods

A cross-sectional study was conducted between 2022 and 2023 by members of the 22nd Chinese medical team (China-Uganda Friendship Hospital (Naguru)) to Uganda. An online questionnaire was distributed to overseas Chinese workforce residing in Uganda, with guidance provided to ensure accurate completion. The survey collected data on socio-demographic characteristics, malaria infection and treatment status, malaria prevention knowledge, attitudes, and practices, access to preventive materials and measures, and prevention needs. A total of 798 valid responses were obtained, covering participants from 32 districts across Uganda, including capital city Kampala. Data were analyzed using descriptive statistics, bivariate analyses, and multivariable logistic regression. Multivariable logistic regression was used to estimate adjusted odds ratios, and a two-sided P value < 0.05 was considered statistically significant.

Results

Malaria infection was defined as self-reported history of malaria during the stay in Uganda. A total of 798 valid online questionnaires were collected and analyzed, of which 44.5% (n = 355) of respondents reported a history of malaria infection. Multivariable logistic regression analysis revealed that being male (AOR = 1.92, 95% CI: 1.27–2.91), having a higher number of visits to Uganda (AOR = 1.28, 95% CI: 1.06–1.53), longer duration of work and residence in Uganda (AOR = 1.38, 95% CI: 1.19–1.61), negative attitudes on malaria prevention (AOR = 2.49, 95% CI: 1.41–4.40), higher mosquito biting frequency (AOR = 1.31, 95% CI: 1.10–1.56), having a higher presence of malaria cases in the surrounding population (AOR = 1.95, 95% CI: 1.51–2.51), and access to antimalarial drug supplies (AOR = 1.48, 95% CI: 1.04–2.11) were significantly associated with an increased likelihood of malaria infection. In contrast, respondents with a bachelor’s degree or higher education level (AOR = 0.65, 95% CI: 0.44–0.97) and those living with a spouse (AOR = 0.57, 95% CI: 0.36–0.91) had a lower likelihood of malaria infection.

Conclusion

This study revealed a substantial burden of self-reported malaria infection among Chinese workforce in Uganda, with recurrent infections underscoring the significant health threats faced by this population. To effectively reduce the malaria burden, an integrated malaria control strategy is required. The strategy should include improving employees’ living conditions and access to healthcare, optimizing work and residential environments to reduce mosquito biting frequency, and implementing systematic and continuous health education programs to strengthen malaria-related knowledge, preventive awareness, and self-efficacy. Establishing a standardized and sustainable supply system for preventive materials and antimalarial drugs is also critical. Comprehensive implementation of these measures will not only protect the health of overseas employees and support stable business operations, but also contribute to sustaining China’s malaria-free status and provide valuable guidance for global malaria control efforts.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-026-26645-5.

Introduction

Malaria remains one of the world’s three major public health challenges, alongside HIV/AIDS and tuberculosis, as recognized by the World Health Organization (WHO) [1]. According to WHO statistics, an estimated 263 million malaria cases and 597,000 deaths occurred worldwide in 2023, with the vast majority of cases and fatalities concentrated in sub-Saharan Africa [1, 2]. Uganda is designated by the WHO as one of the “High Burden to High Impact” (HBHI) countries, experiencing persistently high transmission intensity and widespread population exposure. Despite decades of control efforts, malaria remains a leading cause of morbidity in Uganda, posing ongoing challenges for both local residents and visitors [1, 3].

As China’s international engagement has expanded under the Belt and Road Initiative, a growing number of Chinese enterprises and professionals have been stationed in sub-Saharan Africa, including Uganda. Data from China’s Ministry of Commerce indicates that from January to August 2025, a total of 271,000 Chinese workers were dispatched for overseas labor cooperation [4]. These expatriates engage in diverse occupational settings such as construction, engineering, manufacturing, commerce, education, and cultural exchange. The breadth of these work environments contributes to substantial heterogeneity in their potential exposure to malaria.

Notably, after years of sustained effort, China was certified as malaria-free by the WHO on June 30, 2021, becoming the first country in the WHO Western Pacific Region to receive this certification in over 30 years [5, 6]. This achievement signifies the successful interruption of local malaria transmission in China. However, malaria remains highly endemic in Uganda, and imported malaria continues to pose a persistent challenge for malaria-eliminated countries [3]. Overseas Chinese workers stationed in Uganda represent a mobile population with limited prior exposure to malaria, and their occupational and living environments may place them at increased likelihood of infection if preventive measures are insufficient.

While malaria prevention and control among local Ugandan populations have been extensively studied, evidence remains limited regarding malaria infection history, prevention practices, and associated factors among the overseas Chinese workforce in Uganda. Addressing this gap is important for designing targeted interventions for expatriate populations and for strengthening strategies to prevent imported malaria.

Therefore, this study conducted a questionnaire-based survey to collect information on socio-demographic characteristics, self-reported malaria infection history and treatment, malaria prevention knowledge, attitudes and practices (KAP), access to preventive resources, and prevention needs among overseas Chinese workers residing in Uganda. The findings aim to identify gaps in malaria prevention in this population and to provide evidence to inform targeted malaria control strategies and improve cross-border malaria prevention efforts.

Methods

Study design and sample

This study was a cross-sectional survey designed to investigate the epidemiological characteristics of malaria, the current status of prevention and control, and the factors for malaria infection among overseas Chinese workforce in Uganda. The survey was conducted across 32 districts, including capital city Kampala, and was implemented by researchers from the 22nd Chinese medical team to Uganda (China-Uganda Friendship Hospital (Naguru)). Data collection was conducted between 2022 and 2023. The study population consisted of overseas Chinese workers in Uganda, including employees from Chinese enterprises and organizations across multiple sectors such as construction, engineering, manufacturing, commerce, and education, as well as Chinese freelance workers. Inclusion criteria were: Chinese nationality, ability to read Chinese, work experience in Uganda for more than three months, and installation and use of the WeChat application on a smartphone.

The survey link was distributed through targeted channels, including the Chinese Chamber of Commerce in Uganda and medical networks. To improve accessibility and reduce potential selection bias, researchers provided on-site assistance for participants with limited internet access or lower digital literacy. Assistance included explaining questionnaire items when needed and supporting completion through paper-based or interviewer-assisted entry, while responses were recorded according to participants’ own answers.

The sample size was determined based on the planned multivariable logistic regression analysis. We used G*Power to estimate the required sample size for this study. Based on literature review and the requirements of multivariable analysis, the parameters were set as follows: α = 0.05, power (1 − β) = 0.95, and effect size f² = 0.15. According to the study design, the total degrees of freedom (df) for the predictors was 24. By entering these parameters into the software, the estimated total sample size was 238. Preliminary data collected prior to the formal survey indicated that the malaria infection rate exceeded 40%. To ensure model stability and to meet the empirical criterion of EPV ≥ 10, the theoretical lower bound obtained from G*Power was further adjusted based on the number of events per variable. With df = 24, a total of 240 events were required, corresponding to an overall sample size of N = 240 / 0.40 = 600. Considering an anticipated non-response rate of approximately 20% in the online survey, the minimum required sample size was determined to be 750 participants. Ultimately, 798 complete and valid questionnaires were collected [7, 8]. The survey sample covers workers from various sectors (e.g., construction, business, education), ensuring that the results reflect the diversity of the Chinese workforce in Uganda.

Measuring tools

The questionnaire consisted of three main Sect [9]. The first section collected socio-demographic characteristics, including gender, age, number of visits to Uganda, duration of work and residence in Uganda, educational level, professional title, occupation, living arrangement, and location of work or residence. The second section focused on malaria infection and treatment history, covering the number of malaria episodes, the level of concern after infection, common symptoms experienced during infection, the treatment options used, and the preferred healthcare facilities for seeking treatment. The third section assessed participants’ knowledge, attitudes, and practices (KAP) regarding malaria prevention and control. It included questions on malaria-related knowledge, preventive attitudes and behaviors, accessibility of preventive supplies, and perceived needs for external support in malaria prevention.

The outcome variable, malaria infection, was defined as a self-reported history of malaria diagnosis during the respondent’s stay in Uganda, based on questionnaire responses. Most reported malaria episodes were diagnosed by healthcare professionals at medical facilities, including local hospitals and the Chinese medical teams. In routine practice, diagnostic testing typically follows Uganda’s national malaria guidelines, which recommend parasitological confirmation using malaria rapid diagnostic tests (RDTs) [10]. Although this approach may be subject to recall bias, it is widely used in epidemiological surveys among mobile populations where standardized diagnostic records are often unavailable. Further education was defined as post-secondary education other than university degrees, including vocational or technical training.

Statistical analysis

All statistical analyses were performed using SPSS version 26.0. Descriptive statistics were used to characterize the study population: categorical variables are presented as frequencies (percentages) and continuous variables as mean ± standard deviation. Bivariate analyses were conducted using chi-square tests or Fisher’s exact tests (when > 20% of cells had expected counts < 5). Variables with P < 0.20 in bivariate analyses, as well as variables considered epidemiologically important based on prior literature and the study objectives, were included as candidate variables in the multivariable binary logistic regression model. A forward stepwise selection procedure (entry criterion: P < 0.05; removal criterion: P < 0.10) was applied to identify factors independently associated with malaria infection. This approach was used to achieve a parsimonious model while minimizing the risk of overfitting, given the number of candidate variables relative to the number of outcome events. Variables of epidemiological importance were also evaluated during model construction. The final model results are presented as adjusted odds ratios (AORs) with corresponding 95% confidence intervals (CIs). Model fit was assessed using the Hosmer–Lemeshow goodness-of-fit test. A two-sided P value < 0.05 was considered statistically significant in the final multivariable model.

Result

Socio-demographic characteristics

A total of 798 complete and valid questionnaires were collected in this survey. Respondents from 32 out of the 136 districts in Uganda were included, with the majority (70.4%, n = 562) located in the capital, Kampala (The detailed geographic distribution of Chinese workforce is provided in Supplementary Material 1). Among all respondents, the male-to-female ratio was 3.16:1. The mean age was 38.6 (SD ± 9.34) years, aged 20–66 years (100%). The largest proportion of participants (41.6%, n = 332) were aged 30–39 years. Additionally, 17.0% (n = 136) of respondents co-resided with their spouse, while the majority (83.0%, n = 662) had other living arrangements, such as living alone or with children, parents, or other relatives. The education level of the respondents was predominantly bachelor’s degree (40.6%, n = 324), followed by Secondary or below (30.7%, n = 245). Those with a postgraduate degree also constituted a considerable proportion (10.5%, n = 84). Among the occupation, construction, engineering, and manufacturing accounted for 57.5%, business and services 15.5%, education, management, and clerical services 6.4%, and other 17.4%. Over half of the participants (53.0%, n = 423) mentioned having no professional title or unknown. Nearly 55.3% of participants had more than three years of work or life experience in Uganda, and a significant proportion (49.4%, n = 394) had been visited to this country three or more times (Table 1).

Table 1.

Socio-demographic characteristics of Chinese workforce (n = 798)

Variable	Category	Frequency	Percentage
Gender	Male	606	75.9%
	Female	192	24.1%
Age group (years)	20–29	139	17.5%
	30–39	332	41.6%
	40–49	187	23.4%
	≥ 50	140	17.5%
Duration of stay (years)	≤ 1	259	32.5%
	2	98	12.3%
	3	75	9.4%
	≥ 4	366	45.9%
Uganda visits(times)	1	256	32.1%
	2	148	18.5%
	3–4	285	35.7%
	≥ 5	109	13.7%
Educational level	Secondary or below	245	30.7%
	Further education	145	18.2%
	Bachelor’s degree	324	40.6%
	Postgraduate degree	84	10.5%
Professional rank	Junior	111	13.9%
	Intermediate	170	21.3%
	Senior	94	11.8%
	No title or unknown	423	53.0%
Occupation	Construction and Engineering	320	40.1%
	Manufacturing	139	17.4%
	Business and services	124	15.5%
	Educational and clerical services	51	6.4%
	Other	164	20.6%
Family living arrangements	With spouse	136	17.0%
	Other living arrangements	662	83.0%
Location of work or residence	Urban	466	58.4%
	Field camp	196	24.6%
	Industrial area	136	17.0%

Other living arrangements included: living alone, or with children, parents, or other relatives

The profile of housing and malaria-preventive infrastructure in different zones

Survey results indicated that 24.6% (n = 196) of respondents resided in field camps (Table 1), primarily occupying light steel-structured prefabricated housing or mixed brick-and-concrete structures (Fig. 1A-E). These structures were simply designed but equipped with screened windows and insecticide-treated door curtains. The living quarters were routinely equipped with core malaria prevention amenities, including insecticide-treated bed nets (ITNs), indoor residual spraying (IRS), and insecticide aerosol sprays. Additionally, the compound was equipped with drainage systems, and ground surfaces were paved with cement or brick. The entire premises were enclosed by perimeter fencing, forming a physical barrier that isolated them from the surrounding environment. Furthermore, the residential structures were designed to optimize ventilation and natural lighting, thereby promoting indoor air circulation and dryness to reduce potential mosquito breeding conditions.

Fig. 1.

Residential settings of Chinese workforce in Uganda. A-E Key features of the field camp environment: (A, B) Temporary housing buildings in field camps and surrounding hardened sites; (C) Representative indoor mosquito prevention measures, including a long-lasting insecticidal net (LLIN), an electric mosquito swatter, and window mesh screens; (D) Open concrete drainage channel for surface runoff; (E) Designated buffer zone separating the living area from the peripheral environment. F, G Standardized single-family and modern multi-story residential building in an urban setting.

In contrast, 466 (58.4%) respondents lived in permanent buildings located in urban areas (Table 1). Two common residential types were identified (Fig. 1F-G): (1) multi-story apartment complexes, typically constructed with reinforced concrete and characterized by high resident density; and (2) Detached houses, which featured lower building density, more spacious private yards, and relative higher vegetative coverage (The detailed content of the figure is provided in Supplementary Material 2).

Malaria infection and treatment status

Of the 798 respondents, 355 (44.5%) reported a history of malaria infection during their stay in Uganda. Regarding common symptoms following malaria infection, the multiple-response data revealed that fever (33.7%, n = 269), body aches (32.1%, n = 256), chills (24.2%, n = 193), dizziness (with mental confusion) (24.1%, n = 192) were the most frequently reported. Additionally, local hospitals and Chinese medical teams (China-Uganda Friendship Hospital (Naguru)) were the most common choices for seeking treatment. Among respondents with malaria infection, 63 received oral medications and 112 received intravenous therapy, with respective proportions of 17.7% and 31.5%, a further 115 (32.4%) received both oral medication and intravenous therapy. Those who required hospitalization represented 15.5% of all infected respondents, while the remaining 2.8% (n = 10) received no treatment or relied on traditional Chinese herbal medicine. Beyond clinical and therapeutic aspects, the level of concern regarding malaria infection was also investigated among the participants. The majority (77.7%, n = 532) reported a moderate level of worry, whereas a minimal proportion (2.8%, n = 22) expressed no concern. (Table 2).

Table 2.

Characteristics of malaria infection and treatment among Chinese workforce (n = 798)

Variable	Category	Frequency	Percentage
Number of malaria infections	0	443	55.5%
	1	118	14.8%
	1–5	163	20.4%
	5–10	56	7.0%
	10–20	15	1.9%
	>20	3	0.4%
Level of malaria concern	High	244	30.6%
	Moderate	532	66.7%
	Low	22	2.8%
Symptom	Fever	269	33.7%
	Body aches	256	32.1%
	Headache	156	19.5%
	Chills	193	24.2%
	Dizziness (with mental confusion)	87	11.0%
	Presence of mild or atypical symptoms	108	13.5%
Treatment modality(n = 355)	Oral medication	63	17.7%
	Intravenous therapy	112	31.5%
	Combined oral and intravenous therapy	115	32.4%
	Hospitalization	55	15.5%
	Other (no treatment or traditional herbal)	10	2.8%
Priority-choice healthcare institution	Local hospital	491	61.5%
	Chinese Medical Team	318	39.8%
	On site medical personnel	172	21.6%
	Self-treatment	94	11.8%

Percentages may not sum to 100% due to rounding. Both symptoms and Priority-choice healthcare institution are multiple-response questions, for which participants could select more than one option. Percentages were calculated as the number of respondents selecting each option divided by the total sample size

Knowledge, Attitudes, and Practices (KAP) regarding malaria prevention and control

A total of 31.8% of respondents reported being very familiar with malaria-related knowledge, 51.5% indicated having a moderate level of understanding, and the remaining 16.7% reported only limited awareness. With regard to the attitude of malaria prevention and control, 33.8% of respondents held positive attitudes, a further 21.8% mentioned conditional acceptance, stating that control measures were acceptable only if they did not interfere with work or daily life. In contrast, 10.9% held negative attitudes, while the remaining 33.5% held ambivalent attitudes. Regarding external support for malaria control, 63.5% (n = 507) of respondents expressed a strong need, 31.1% (n = 248) reported a moderate need, while 5.4% (n = 26) reported no need for such assistance. Additionally, in response to a multiple-response question regarding the most concerning infectious diseases, Ebola hemorrhagic fever (EHF) was the most frequently cited concern (85.6%, n = 683), followed by malaria (61.7%, n = 492) and HIV/AIDS (55.5%, n = 443). Cholera or typhoid fever and COVID-19 were mentioned by 30.0% (n = 310) and 25.0% (n = 217) of participants, respectively, while other diseases accounted for 34.3% (n = 274) (Table 3).

Table 3.

Characteristics of malaria knowledge, attitudes, practices, and perceived needs (n = 798)

Variable	Category	Frequency	Percentage
Attitudes on malaria prevention
	Positive	270	33.8%
	Conditional acceptance	174	21.8%
	Negative	87	10.9%
	Ambivalent	267	33.5%
Malaria knowledge level
	High	254	31.8%
	Moderate	411	51.5%
	Low	133	16.7%
The need for external support
	High	507	63.5%
	Moderate	248	31.1%
	None	26	5.4%
Most concerning infectious disease
	Ebola hemorrhagic fever (EHF)	683	85.6%
	Malaria	492	61.7%
	HIV/AIDS	443	55.5%
	Cholera/Typhoid	310	38.8%
	COVID-19	217	27.2%
	Other	274	34.3%
Frequency of mosquito bites
	Nearly none	51	6.4%
	Low (occasional)	387	48.5%
	Moderate (several times per week)	152	19.0%
	High (≥ 1/daily)	208	26.1%
Environmental management frequency
	Irregular	116	14.5%
	Infrequent (≈ biannually)	72	9.0%
	Occasional (every 2–3 months)	184	23.1%
	Frequent (≥ 1/monthly)	426	53.4%
Peak-hour exposure to high-exposure habitats
	Seldom or never	431	54.0%
	Occasionally	286	35.8%
	Frequent	81	10.2%
Frequency of contact with local staff
	Occasional	73	9.1%
	About 7 days per month	6	0.8%
	3–4 days per week	26	3.3%
	Almost daily	693	86.8%
Malaria cases in the surrounding population
	Nearly none	51	6.4%
	Low (occasional cases)	524	65.7%
	Moderate (≥ 1 cases/month)	172	21.6%
	High (≥ 1 case/week)	51	6.4%
Malaria training access
	Yes	457	57.3%
	No	332	41.6%
	Regular training	9	1.1%
On site medical personnel
	Yes	216	27.1%
	No	582	72.9%
Availability of antimalarial drugs
	Yes	485	60.8%
	No	256	32.1%
	Unknown	57	7.1%

Percentages may not sum to 100% due to rounding. Most concerning infectious disease is multiple-response questions, for which participants could select more than one option. Percentages were calculated as the number of respondents selecting each option divided by the total sample size

Mosquito bites were highly frequent among Chinese workforce in Uganda. Survey data indicated that 26.1% (n = 208) of respondents experienced a high frequency of mosquito bites (≥ 1 per day), 19.0% (n = 152) reported a moderate frequency (several times per week), 48.5% (n = 387) reported a low frequency (occasional bites), while 6.4% (n = 51) reported nearly no mosquito bites. Despite the implementation of related interventions, substantial individual variation persisted in preventive behaviors. The total 426 respondents (53.4%) reported conducting regular environmental management (e.g., weed removal, stagnant water elimination, and insecticide spraying) at least monthly. In comparison, 116 individuals (14.5%) applied these measures irregularly. With regard to exposure to environments with higher mosquito density during peak mosquito activity periods (early morning and evening), 81 respondents (10.2%) mentioned frequently visiting wetlands, riversides, or grasslands; 286 (35.8%) reported occasional visits; and 431 (54.0%) seldom or never entered such area. Most respondents (86.8%, n = 693) reported almost daily contact with local staff, the remainder reported contact 3–4 days per week (3.3%, n = 26), approximately 7 days per month (0.8%, n = 6), or occasionally (9.1%, n = 73). In addition, regarding the presence of malaria cases among the surrounding population, the largest proportion of respondents (65.7%, n = 524) reported occasional malaria case, followed by those who had contact with more than 1 case per month, accounting for 21.6% (n = 172). The remaining respondents included those with contact with more than 1 case per week and those who almost never had contact, both groups comprised 51 individuals, each representing 6.4% (Table 3).

Availability of malaria healthcare resources

Survey results indicated that antimalarial drugs were available at the workplaces of 485 respondents (60.8%), the proportion who reported having on-site medical personnel at their work or residential locations was 27.1% (n = 216). Furthermore, while 57.3% (n = 457) of participants had received malaria-related training, regular training coverage was markedly low, reported by only 9 individuals (1.1% of the total sample) (Table 3).

Bivariate analyses identified multiple variables potentially associated with malaria infection (Tables 4 and 5). Variables meeting the screening criterion (P < 0.20) and those deemed epidemiologically relevant were entered into the multivariable logistic regression model. Forward stepwise selection was then applied to derive a parsimonious final model, which retained nine independent predictors. Ordinal variables (e.g., duration of stay in Uganda and number of visits to Uganda) were analyzed as continuous or ordinal, while nominal multi-category variables (e.g., living arrangements and malaria prevention knowledge) were dummy-coded.

Table 4.

Bivariate analysis of sociodemographic factors associated with malaria infection (n = 798)

Variable	Category	Self-reported malaria history		Chi-square value(χ2)	P-value
		Yes(%)	No(%)
Gender	Male	85.1	68.6	29.178	0.000
	Female	14.9	31.4
Age group (years)	20–29	11.8	21.9	26.350	0.000
	30–39	38.6	44.0
	40–49	26.8	20.8
	≥ 50	22.8	13.3
Duration of stay (years)	≤ 1	21.1	44.4	73.187	0.000
	2	8.3	14.8
	3	11.1	7.6
	≥ 4	59.4	33.1
Uganda visits(times)	1	22.3	40.0	55.137	0.000
	2–3	14.6	21.7
	3–4	42.5	30.2
	≥ 5	20.6	8.1
Educational level	Secondary or below	37.7	25.1	26.336	0.000
	Further education	18.9	17.6
	Bachelor’s degree	37.7	42.9
	Postgraduate degree	5.6	14.4
Professional rank	Junior	14.4	13.5	0.971	0.808
	Intermediate	20.0	22.3
	Senior	11.3	12.2
	No title or unknown	54.4	51.9
Occupation	Architecture and engineering	45.6	35.7	31.891	0.000
	Manufacturing	21.7	14.0
	Business and services	13.8	16.9
	Educational and clerical services	2.8	9.3
	Other	16.1	24.2
Family living arrangements	With spouse	14.1	19.4	3.958	0.047
	Other living arrangements	85.9	80.6

Table 5.

Bivariate analysis of associations between malaria infection and related knowledge, attitudes, practices, and perceived needs (n = 798)

Variable	Category	Self-reported malaria history		Chi-square value(χ2)	P-value
		Yes(%)	No(%)
Attitudes on malaria prevention	Positive	28.2	38.4	22.593	0.000
	Conditional acceptance	23.4	20.5
	Negative	16.1	6.8
	Ambivalent	32.4	34.3
Malaria knowledge level	High	27.9	35.0	4.594	0.101
	Moderate	54.6	49.0
	Low	17.5	16.0
The need for external support	High	59.7	66.6	6.638	0.036
	Moderate	33.0	29.6
	None	7.3	3.8
Frequency of mosquito bites	Nearly none	3.1	9.0	30.555	0.000
	Low (occasional)	43.9	52.1
	Moderate (several times per week)	18.6	19.4
	High (≥ 1/daily)	34.4	19.4
Environmental management frequency	Infrequent (≈ biannually)	10.7	7.7	5.413	0.144
	Occasional (every 2–3 months)	19.7	25.7
	Frequent (≥ 1/monthly)	54.4	52.6
	Irregular	15.2	14.0
Peak-hour exposure to high-exposure habitats	Seldom or never	47.9	58.9	13.195	0.001
	Occasionally	38.3	33.9
	Frequent	13.8	7.2
Frequency of contact with local staff	Occasional	7.0	10.8	10.947	0.009
	About 7 days per month	0.0	1.4
	3–4 days per week	2.3	4.1
	Almost daily	90.7	83.7
Malaria cases in the surrounding population	Nearly none	1.4	10.4	59.591	0.000
	Low (occasional cases)	59.2	70.9
	Moderate (≥ 1	29.9	14.9
	High (≥ 1 case/week)	9.6	3.8
Malaria training access	Yes	55.5	58.7	4.524	0.104
	No	42.5	40.9
	Regular training	2.0	0.5
On site medical personnel	Yes	31.0	23.9	4.973	0.026
	No	69.0	76.1
Availability of antimalarial drugs	Yes	66.2	56.4	14.689	0.001
	No	30.1	33.6
	Unknown	3.7	9.9

Multivariable logistic regression analysis indicated that males had significantly increased odds of infection than females (AOR = 1.92, 95% CI: 1.27–2.91). In terms of educational level, respondents with a bachelor’s degree or a postgraduate degree had lower odds of malaria infection compared to those with secondary education or less, with adjusted odds ratios of 0.65 (95% CI: 0.44–0.97) and 0.41 (95% CI: 0.22–0.77), respectively. Regarding living arrangements, the odds of malaria infection are decreased by 43.0% among respondents co-residing with their spouse as compared to those with other living arrangements (AOR = 0.57, 95% CI: 0.36–0.91). Additionally, a longer duration of stay in Uganda (AOR = 1.38, 95% CI: 1.19–1.61) and a greater number of visits to the country (AOR = 1.28, 95% CI: 1.06–1.53) were both significantly associated with increased odds of infection. Similarly, a higher number of malaria cases in the surrounding population (AOR = 1.95, 95% CI: 1.51–2.51) and a higher frequency of mosquito bites (AOR = 1.31, 95% CI: 1.10–1.56) were both significantly associated with an increased likelihood of malaria infection. Respondents with negative attitudes toward malaria control had 2.49-fold higher odds of malaria infection relative to those with positive attitudes (AOR = 2.49, 95% CI: 1.41–4.40). Notably, individuals at workplaces providing antimalarial drugs demonstrated higher infection odds (AOR = 1.48, 95% CI: 1.04–2.11) than those at workplaces without such resources. The Hosmer–Lemeshow test indicated an acceptable model fit (χ2 = 7.572, df = 8, P = 0.476) (Table 6).

Table 6.

Multivariable logistic regression analysis of malaria infection in Chinese workforce in Uganda

Variable	Category	Standard error	P-value	AOR (95% CI)
Gender	Female			1
	Male	0.211	0.002	1.92 (1.27–2.91)
Educational level	Secondary or below			1
	Further education	0.236	0.357	0.80 (0.51–1.28)
	Bachelor’s degree	0.201	0.033	0.65 (0.44–0.97)
	Postgraduate degree	0.323	0.006	0.41 (0.22–0.77)
Duration of stay		0.077	0.000	1.38 (1.19–1.61)
Uganda visits		0.093	0.009	1.28 (1.06–1.53)
Family living arrangements	Other living arrangements			1
	With spouse	0.238	0.018	0.57 (0.36–0.91)
Attitudes on malaria prevention	Positive			1
	Conditional acceptance	0.226	0.134	1.40 (0.90–2.19)
	Negative	0.291	0.002	2.49 (1.41–4.40)
	Ambivalent	0.201	0.328	1.22 (0.82–1.81)
Malaria cases in the surrounding population		0.129	0.000	1.95 (1.51–2.51)
Frequency of mosquito bites		0.089	0.002	1.31 (1.10–1.56)
Availability of antimalarial drugs	No			1
	Yes	0.182	0.031	1.48 (1.04–2.11)
	Unclear	0.376	0.241	0.64 (0.31–1.34)

Forward stepwise binary logistic regression analysis indicating the net effect of the explanatory variables on the outcome variables

Discussion

Uganda, a country in sub-Saharan Africa, faces constraints in its public health infrastructure, characterized by limited medical facilities and an uneven distribution of healthcare resources [11]. Moreover, 41.6% of Chinese workforce in Uganda worked or lived in remote field camps or industrial zones far from urban areas, where personnel were widely dispersed, and only 27.1% of employees were equipped with on-site medical staff. Systematically characterising malaria epidemiology and current control practices among Chinese workforce in Uganda is therefore critical for developing targeted interventions and strengthening prevention capacity in this population. Evidence from such work will also inform strategies to prevent imported malaria, help consolidate elimination gains in malaria-free countries, and contribute to global malaria control and elimination efforts.

This study found that male respondents had a significantly higher likelihood of malaria infection than females (AOR = 1.92, 95% CI: 1.27–2.91), a finding consistent with previous reports [12, 13]. Several factors may explain this difference. First, studies have suggested that males exhibit lower immune response levels compared with females, which may be related to the modulatory effects of sex hormones on host immune regulation [14]. Second, the survey revealed that male employees were more likely to engage in outdoor activities both during and outside working hours, while their adherence to mosquito prevention measures was generally lower than that of females, thereby increasing their exposure to mosquito bites.

This study found that participants with a bachelor’s degree or higher had a significantly lower likelihood of malaria infection in multivariable analysis (AOR = 0.65, 95% CI: 0.44–0.97), whereas those with an further education showed no significant difference compared with individuals who had a secondary education or below (P > 0.05). Individuals with higher educational attainment generally possess better health literacy, stronger preventive behaviors such as consistent use of insecticide-treated nets and timely healthcare-seeking, and tend to enjoy superior living conditions and higher socioeconomic status, all of which collectively contribute to reduced infection likelihood [15]. Meanwhile, in this study population, participants with further education were more often engaged in technical or field-based positions and typically resided in collective dormitories. Such occupational outdoor exposure and shared housing environments may offset the behavioral and cognitive advantages associated with education, resulting in a comparable likelihood to those with lower educational backgrounds. Previous studies have similarly demonstrated that housing quality, occupational exposure, and health behaviors play important mediating roles in the relationship between education and malaria infection [16, 17].

The results of this study showed that employees living with their spouses had a significantly lower likelihood of malaria infection compared with those under other living arrangements (AOR = 0.57, 95% CI: 0.36–0.91). This finding can be explained by several interrelated mechanisms. First, living with a spouse often provides a family environment with stronger emotional support, more stable daily routines, and greater sharing of preventive resources such as bed nets, thereby reducing opportunities for mosquito bites and parasite transmission [18]. Second, co-residence with parents or relatives usually implies larger household sizes, limited living space, and lower availability or sharing efficiency of mosquito prevention resources such as bed nets or repellents [19]. Moreover, if household members have a history of malaria, intra-household transmission chains may occur. Third, although individuals living alone are fewer in number, they may lack family supervision, have lower adherence to preventive practices, all of which increase exposure likelihood. In addition, from both psychological and physiological perspectives, living alone is often associated with higher levels of psychological stress, loneliness, and other negative emotional states. Previous studies have suggested that such chronic psychosocial stress can disrupt the neuroendocrine–immune axis, resulting in impaired immune function and heightened susceptibility to diseases [20]. These mechanisms suggest that living arrangements reflect not only physical environmental conditions but also social support structures, preventive behaviors, and intra-household transmission dynamics, all of which substantially influence malaria infection. Therefore, malaria prevention programs for overseas Chinese workforce should incorporate living arrangements into the assessment of malaria infection likelihood and develop targeted control strategies for those living alone or with parents or relatives.

Survey findings indicated that respondents with negative attitudes toward malaria prevention had a higher likelihood of infection (AOR = 2.49, 95% CI: 1.41–4.40). This suggests that confidence in prevention is not merely a subjective attitude but may also reflect behavioral tendencies and a sense of health responsibility. Employees with positive preventive cognitions generally demonstrated higher adherence to protective practices in daily life, such as consistent use of bed nets and repellents, as well as reducing outdoor activities, thereby objectively lowering their likelihood of infection. Conversely, individuals lacking such beliefs may develop a “fatalistic” mindset after experiencing recurrent infections themselves or occasionally witnessing the ineffectiveness of others’ preventive efforts, leading to skepticism, complacency, or even abandonment of protective behaviors and creating a negative cycle of cognition, behavior, and outcomes [21]. Moreover, the self-efficacy represented by this variable constitutes an important psychological foundation for health behavior change [22]. These results underscore that malaria prevention should not remain limited to knowledge dissemination but must also aim to strengthen individuals’ confidence in their ability to manage their own health, thereby fostering proactive protective awareness and enabling more effective and sustainable behavioral interventions.

In this study, data indicated that 86.8% of Chinese workforce in Uganda reported almost daily contact with local Ugandan employees. However, multivariable analysis revealed that frequent contact with local Ugandan employees was not statistically associated with malaria infection at the individual level (P > 0.05). This finding suggests that routine interpersonal contact alone is not decisively associated with an increased likelihood of malaria infection. Therefore, Chinese workforce in Uganda do not need to be overly concerned about regular interactions with local colleagues during daily work and communication. In contrast, another variable—presence of malaria cases in the surrounding population—was significantly associated with individual infection (AOR = 1.95, 95% CI: 1.51–2.51). This association is likely not attributable to frequent social contact per se, but rather to prolonged residence in areas with high densities of infected individuals (e.g., camps or communities). Malaria transmission depends on the human–mosquito–human cycle; the more infected individuals present in the environment, the higher the likelihood of being bitten by Anopheles mosquitoes, thereby increasing the likelihood of individual infection [23, 24]. These findings suggest that the key determinant of malaria infection lies not in the frequency of human contact, but in the density of infected individuals within one’s surrounding environment. Accordingly, malaria prevention strategies in Chinese enterprises should prioritize the identification and management of regions with high infection rates, shifting the focus from merely limiting interpersonal contact to reducing the density of infection sources and minimizing environmental exposure.

Multivariable logistic regression analysis demonstrated a significant association between mosquito biting during residence in Uganda and the likelihood of malaria infection (AOR = 1.31, 95% CI: 1.10–1.56), with higher biting frequency corresponding to greater infection likelihood. This finding is logically consistent with the transmission mechanism of malaria, as the frequency of mosquito bites directly determines the level of mosquito exposure and represents a central link in the transmission chain [25]. At the same time, the frequency of mosquito bites can serve as an indirect indicator of local mosquito density, the adequacy of vector control measures, and individual adherence to protective practices. Frequent bites may not only reflect poor environmental sanitation or limited access to protective resources but may also signal inadequate personal protective behaviors. Thus, bite frequency serves not only as a measure of mosquito exposure but also as an integrated marker of behavioral and environmental exposure. It should therefore be considered an important criterion for identifying high-exposure populations and optimizing prevention strategies, including enhanced entomological surveillance and targeted training on personal protection in priority areas, to mitigate transmission from the outset.

In this study, employees with access to antimalarial drug supplies had a higher likelihood of malaria infection (AOR = 1.48, 95% CI: 1.04–2.11), which may reflect a reverse causality pattern. Antimalarial drugs were stocked because many employees worked or resided in areas with high malaria transmission risk. Moreover, easy access to drugs could reduce adherence to preventive measures such as regular use of insecticide-treated nets or avoidance of outdoor exposure. Inadequate drug quality, self-medication, and delayed treatment may also weaken the potential benefits of drug availability [26]. Although prompt treatment can interrupt transmission by reducing the human infectious reservoir, drug storage alone does not ensure effective malaria control. A comprehensive workplace health protection system that integrates vector control, early diagnosis, environmental management, and health education is essential to translate drug accessibility into actual disease prevention [27, 28].

Environmental interventions have long been regarded as a cornerstone of malaria control, theoretically aiming to reduce infection likelihood by eliminating mosquito breeding sites and interrupting transmission pathways [29]. However, in this study, no significant association was observed between routine environmental management measures (such as grass cutting and water drainage) and malaria infection (P > 0.05). Several factors may explain this finding. Most participants resided in well-constructed housing equipped with cement floors, screened windows, closed doors, and adequate drainage, all of which substantially reduced opportunities for mosquito breeding [30]. In addition, widespread use of mosquito nets, repellents, and insecticide sprays further diminished the relative contribution of environmental measures. Notably, previous studies have indicated that unsupervised environmental interventions may paradoxically increase exposure, as individuals entering wetlands or grassy areas with high mosquito density may experience more bites [31]. This was also evident in our survey: 10.2% of respondents reported frequently visiting wetlands or riverbanks in early mornings or evenings, and such behavior was associated with infection (P < 0.05). These findings underscore that environmental management in malaria control should avoid a “one-size-fits-all” approach, and instead be tailored to local ecological conditions and residential characteristics.

Conclusion

Among the Chinese workforce in Uganda, a high proportion of individuals reported having experienced malaria, which not only disrupts normal business operations but also seriously impacts employees’ physical and mental health. This study suggested several factors significantly associated with malaria infection, including gender, educational level, duration of work and residence in Uganda, number of visits to Uganda, living arrangements, attitudes on malaria prevention, presence of malaria cases in the surrounding population, mosquito biting frequency, and access to antimalarial supplies. Based on these findings, a comprehensive malaria prevention strategy should be implemented, encompassing improvements in living conditions and healthcare access, systematic health education to enhance protective capacity and beliefs, and the establishment of a routine supply mechanism for antimalarial supplies, etc. These measures would not only protect the health of overseas Chinese workforce but also help sustain China’ s malaria-free status and provide guidance for global malaria control efforts.

Limitations

This study has several limitations that should be acknowledged. First, due to the cross-sectional study design, causal relationships between associated factors and malaria infection cannot be established. Second, malaria infection was assessed based on self-reported past diagnosis, which may be subject to recall bias or misclassification. However, most respondents reported malaria episodes diagnosed by healthcare providers, including local hospitals or Chinese medical teams, and the recall period was limited to their stay in Uganda, which may have mitigated recall error. Future studies incorporating laboratory-confirmed diagnoses or medical records are warranted to further improve outcome ascertainment. Third, the study population was restricted to overseas Chinese workforce working in Uganda, which may limit the generalizability of the findings to other expatriate or local populations. In addition, the online questionnaire-based survey design, while improving accessibility and response efficiency, may have introduced selection bias by underrepresenting individuals with limited internet access or lower digital literacy. Fourth, although potential confounding was addressed through multivariable logistic regression and the reporting of adjusted odds ratios, some factors—such as detailed living conditions, mobility patterns, and adherence to preventive measures—were not fully quantified and may have influenced the observed associations. Moreover, formal statistical tests for effect modification were not conducted, and residual confounding cannot be entirely excluded despite careful model adjustment. Given that the outcome was relatively common, the odds ratios derived from logistic regression may overestimate the corresponding risk or prevalence ratios; therefore, these estimates should be interpreted as measures of relative likelihood rather than absolute risk. Finally, future longitudinal and multi-country studies incorporating clinical, behavioral, and entomological data are needed to validate and extend these findings and to further elucidate potential causal pathways.

Authors’ contributions

JJ, GXW and LLM conceived and designed the study. JJ collected the data. JJ, LLM, GXW, and YTC conducted the statistical analysis. JJ, GXW and LLM drafted the manuscript. JJ, GXW, LLM, YTC and ZAJ provided critical feedback and contributed to the revision of the manuscript. All authors reviewed, discussed the results, and endorsed the submission of the final manuscript.

Funding

This study was funded by the In-hospital Research Project of the Third People’s Hospital of Yunnan Province (Project No. 2024SSYKT39).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study received ethical approval from the Ethics Committee of Yunnan Third People’s Hospital (Approval No. 2025KY023). The study was conducted in accordance with the ethical principles of the Declaration of Helsinki. As the research involved an anonymous questionnaire survey with minimal risk to participants, the ethics committee waived the requirement for a written signed consent form. Informed consent was obtained from all participants electronically; a detailed information sheet was presented on the first page of the online questionnaire, explaining the study’s purpose, procedures, and data confidentiality. Proceeding to and submitting the questionnaire was considered implied consent to participate. All methods were performed in accordance with the relevant guidelines and regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.