Effect of cumulative exposure to media channels for malaria messages on knowledge of malaria prevention among women (15–49 years) in Uganda Authors: Charles Natuhamya, Rejoice Uche Obiora, Gideon Ikemdinachi Nwankwo, Delight Mawufemor Agbi, Isaac Isiko, Edson Mwebesa Published: 2026 ABSTRACT IntroductionMalaria remains a leading global public health concern, disproportionately affecting populations in low-resource settings. Uganda continues to contribute substantially to the global malaria burden, yet exposure to malaria-related health messages remains limited. In recent years, diverse media platforms have been adopted to disseminate prevention messages. This study assessed the influence of cumulative exposure to malaria message media channels on knowledge of malaria prevention and its associated factors among women of reproductive age in Uganda.MethodsWe conducted a secondary analysis of the 2018–2019 Uganda Malaria Indicator Survey (UMIS), comprising 7124 women aged 15–49 years selected using a two-stage cluster and stratified sampling design. Knowledge of four prevention methods recommended by World Health Organization (WHO), bed nets, insecticide-treated nets (ITNs), preventive medicine, and indoor residual spraying (IRS) was assessed. Associations between cumulative media exposure and malaria knowledge were examined using t-tests, margins analysis, and mixed-effects negative binomial regression models.ResultsKnowledge of malaria prevention was unevenly distributed: 76.5% of women reported awareness of bed nets, compared with only 9.1% for ITNs, 6.2% for preventive medicine, and 4.3% for IRS. Cumulative exposure to media channels was significantly associated with knowledge of ITNs (p < 0.001), preventive medicine (p = 0.002), and IRS (p < 0.001), but not bed nets. Education, age, wealth, residence, and region were significant determinants of exposure to media channels. Women with secondary or higher education were nearly twice as likely to report exposure to multiple channels compared to uneducated women (Incidence Rate Ratio, IRR = 1.86; 95% Confidence Interval, CI 1.59–2.17). Similarly, women aged ≥ 40 years were 69% more likely (IRR = 1.69; 95% CI 1.35–2.11) to report exposure to multiple channels than those under 20 years. In contrast, rural and refugee women reported significantly lower exposure relative to urban residents (IRR = 0.77 and 0.28, respectively).ConclusionsCumulative multi-channel exposure to malaria prevention messages significantly improves women’s knowledge of ITNs, preventive medicine, and IRS. However, structural inequities in education, wealth, and place of residence limit access to diverse channels. Integrated, context-specific, and equity-focused communication strategies are essential to broaden awareness beyond bed nets and to accelerate progress toward malaria control and elimination in Uganda. Introduction Malaria remains one of the world’s most pressing public health challenges. An estimated 263 million cases and more than 600,000 deaths from malaria were reported globally in 2022, the vast majority occurring in sub-Saharan Africa [1]. The region accounts for over 90% of global malaria morbidity and mortality, underscoring its disproportionate burden on low-resource countries [1]. Uganda is among the most severely affected countries, where malaria is endemic in 96% of its territory and contributes approximately 5% of the global malaria burden [2]. With an estimated incidence of 478 cases per 1000 population, Uganda ranks among the top three countries globally for malaria [3]. Pregnant women and children under five bear the brunt of malaria-related illness and death, despite ongoing prevention campaigns in the country. Over the past two decades, insecticide-treated nets (ITNs), particularly long-lasting insecticidal nets (LLINs), have been at the center of global malaria control strategies [4]. ITNs have demonstrated significant reductions in malaria morbidity and mortality across diverse transmission settings. In 2022, approximately 254 million ITNs were distributed in malaria-endemic countries, with 93% allocated to Sub-Saharan Africa, and Uganda receiving 13.8 million nets [5]. In addition, indoor residual spraying (IRS) has been employed in high-burden districts in Uganda to complement ITN distribution [6]. Despite these interventions, malaria continues to be a leading cause of illness and death in Uganda. For instance, following the 2021 mass ITN distribution campaign, a nationally representative survey was conducted in which 95% of households reported they owned a net, yet only 65% reported net use the night before the survey [7]. This shows that net possession does not automatically translate into consistent use. This possession-use gap reflects not only behavioral choices but also structural and informational constraints; factors such as geographic isolation, distance from health facilities, and traditional housing structures can simultaneously limit exposure to malaria prevention messages and constrain a household’s practical ability to hang or maintain nets. Furthermore, the physical deterioration of nets and perceived ineffectiveness over time further contribute to this discrepancy [8]. Health education and communication have long been recognized as integral components of malaria prevention strategies, within the behavioral change communication umbrella in Sub-Saharan Africa. Traditional and modern media platforms, including radio, television, newspapers, posters, community dramas, and digital channels, have been used to disseminate malaria-related information [9]. Studies from Ghana and Nigeria have shown that mass media exposure has improved awareness and increased ITN usage, particularly among pregnant women [9, 10]. However, the 2019 Uganda Malaria Indicator Survey (MIS) revealed that only 39% of women of reproductive age (15–49 years) had been exposed to malaria messages, with coverage dropping to 16% in refugee settlements [11]. Cumulative exposure to multiple media channels may have a reinforcing effect on knowledge acquisition and adoption of malaria preventive practices. Evidence from a mass media exposure study in Ghana suggests that repeated messaging across diverse outlets not only strengthens recall but also improves the likelihood of translating awareness into consistent preventive actions [12]. Nevertheless, the degree to which cumulative exposure across multiple media platforms enhances knowledge of malaria prevention, particularly among women of reproductive age in Uganda, has not been sufficiently examined. Addressing this evidence gap is crucial for informing targeted health communication strategies that can optimize behavioral outcomes and strengthen malaria control efforts in high-burden communities. This study, therefore, aims to assess the impact of cumulative exposure to media channels on knowledge of malaria prevention in Uganda, focusing on key interventions including bed-net, ITNs, preventive medicine, and IRS. This study was guided by two research questions: (1) Does increasing exposure to malaria message media channels increase the knowledge of malaria prevention methods? and (2) What individual and household factors influence the cumulative exposure to malaria message media channels? Methods Data source and study population The study utilized secondary data based on a two-stage cluster and stratified sampling technique from the Uganda Malaria Indicator Survey (UMIS) of 2018/19 [13]. The study sample consists of 7124 women aged 15–49 who were either permanent residents of the selected households or visitors who stayed in the household the night before the survey. The sampling criteria were previously described in detail [14]. Measurements Exposure to media channels for malaria messages was assessed by asking women if they had seen/heard any messages about malaria six months before the survey and the media channel from which they received the messages. The media channels included: radio, television, poster/billboard, community health worker, community event, interpersonal communication, flyers, social mobilization, social media, and others. Response options were yes/no. The cumulative exposure to media channels for malaria messages, on the other hand, was obtained by summing up the number of media channels a woman reported to have been exposed to, resulting in a count variable. This was the outcome variable used in the regression models. Model covariates included: education level (None, Primary, Secondary or higher), age (Below 20 years, 20–29 years, 30–39 years, 40 or more years), wealth index (Poor, Middle, Rich), place of residence (Urban, Rural, Refugee), religion (Anglican, Catholic, Muslim, Other), and region (Central, East, North, West). Knowledge of malaria prevention was assessed by asking women about the methods for avoiding malaria they knew about, which included: sleeping under a bed net, sleeping under an insecticide-treated net (ITN), taking preventative medication, and indoor residual spraying (IRS), among others. Analyses only included the four listed methods because they are considered by the World Health Organization (WHO) as the key malaria prevention methods [15] and have been found effective in reducing the risk of malaria in Uganda [16]. Statistical analysis Firstly, the data were sorted, and some variables were re-coded. Categorical variables were presented as counts and percentages to describe the study population. Knowledge of each malaria prevention method was presented using proportions disaggregated by women's characteristics, while the effect of the cumulative exposure to media channels for malaria messages on knowledge of malaria prevention methods was assessed using the t-test and margins [17] estimated from mixed-effects logistic regression models. We fitted four models, each including the cumulative exposure to media channels variable as the main predictor. A different malaria knowledge variable was included in each model as the outcome. To determine the individual and household factors influencing cumulative exposure to media channels for malaria messages, mixed-effects negative binomial regression models were specified to account for overdispersion in these data [18], following the rejection of the null hypothesis for the dispersion parameter (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha$$\end{document}α). The mixed-effects models also accounted for cluster variation by including clusters (enumeration areas) as the random effects. Bivariate negative binomial mixed effects models were conducted by including each of the independent variables in the model one at a time, with the cumulative number of exposures to media channels as the outcome. Variables with p ≤ 0.20 were considered for inclusion in the multivariable model [14]. The model was specified as:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$log\left({\mu}_{ij}\right) ={\eta}_{ij}={X}_{ij}\beta +{U}_{j}$$\end{document}logμij=ηij=Xijβ+Ujwhere \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${y}_{ij}$$\end{document}yij is the count outcome for the \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$i$$\end{document}i-th woman in cluster \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$j$$\end{document}j, that follows a Negative Binomial distribution with parameters \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mu}_{ij}$$\end{document}μij and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha$$\end{document}α, where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mu}_{ij}$$\end{document}μij is the expected count, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha$$\end{document}α is the dispersion parameter. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${X}_{ij}\beta$$\end{document}Xijβ are the fixed effects and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${U}_{j}$$\end{document}Uj are the random intercepts for cluster \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$j$$\end{document}j where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${U}_{j} \sim N(0, {\sigma}_{ij}^{2})$$\end{document}Uj∼N(0,σij2). The variance is \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mu}_{ij}+\frac{{\mu}_{ij}^{2}}{\alpha }$$\end{document}μij+μij2α and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$log\left({\mu}_{ij}\right) ={\eta}_{ij}$$\end{document}logμij=ηij is the log link function. Statistical analyses were conducted in Stata 15.0, and the level of statistical significance for all statistical tests was set at p < 0.05. All model assumptions were assessed and verified. Results Characteristics of the study population A total of 7124 women aged 15–49 years were included in the analysis. More than half of the respondents had attained primary-level education (57.1%), while 21.9% had completed secondary education or higher. The majority were aged 20–29 years (52.7%) and resided in rural settings (71.8%). Approximately 57.4% of the women belonged to households classified in the “poor” wealth quintile. With respect to religion, 41.6% identified as Catholic, 31.1% as Anglican, and 10.0% as Muslim. Regional distribution showed that the northern region accounted for the largest proportion of respondents (34.7%), followed by the western (27.4%), eastern (22.3%), and central (15.5%) regions (Table 1).Table 1Women’s demographic characteristicsCharacteristicCountPercent (%)Education level None149621.0 Primary407157.1 Secondary or higher155721.9Age Below 204145.8 20–29375352.7 30–39244634.3 40 and above5117.2Wealth index Poor409057.4 Middle115716.2 Rich187726.4Residence Urban135119.0 Rural511771.8 Refugee6569.2Religion Anglican221331.1 Catholic296141.6 Muslim71410.0 Other123617.4Region Central110515.5 East159122.3 North247534.7 West195327.4 Table 1Women’s demographic characteristicsCharacteristicCountPercent (%)Education level None149621.0 Primary407157.1 Secondary or higher155721.9Age Below 204145.8 20–29375352.7 30–39244634.3 40 and above5117.2Wealth index Poor409057.4 Middle115716.2 Rich187726.4Residence Urban135119.0 Rural511771.8 Refugee6569.2Religion Anglican221331.1 Catholic296141.6 Muslim71410.0 Other123617.4Region Central110515.5 East159122.3 North247534.7 West195327.4 Women’s demographic characteristics Knowledge of malaria prevention methods Overall, 76.5% (95% CI 75.5–77.4) of women identified sleeping under a bed net as a malaria prevention method, while 9.1% (95% CI 8.4–9.8) were knowledgeable about the use of insecticide-treated nets (ITNs). Preventive medicine and indoor residual spraying (IRS) were less frequently reported, at 6.2% (95% CI 5.6–6.7) and 4.3% (95% CI 3.9–4.8), respectively. Stratified analyses revealed statistically significant heterogeneity across sociodemographic strata (Table 2). Knowledge of bed nets varied by education, wealth index, religion, and region (p < 0.05). Knowledge of ITNs, preventive medicine, and IRS demonstrated significant variation across nearly all covariates, with the exception of age. Notably, women with secondary or higher education were substantially more likely to report knowledge of ITNs (16.7%; 95% CI 14.9–18.6) and IRS (10.0%; 95% CI 8.6–11.5) compared to their counterparts with no formal education.Table 2Proportions (95% CI) of women knowledgeable about malaria prevention methods by demographic characteristicsCharacteristicBed net useITN useTaking preventive medicineIRSProportion (95% CI)pProportion (95% CI)pProportion (95% CI)pProportion (95% CI)pOverall76.45 (75.45, 77.42)-9.08 (8.44, 9.77)-6.15 (5.61, 6.73)-4.32 (3.87, 4.82)-Education level None67.25 (64.82, 69.58) < 0.0016.22 (5.10, 7.56) < 0.0017.95 (6.69, 9.44)0.0021.14 (0.71, 1.82) < 0.001 Primary78.95 (77.67, 80.17)7.22 (6.47, 8.06)5.40 (4.75, 6.14)3.34 (2.83, 3.94) Secondary or higher78.74 (76.64, 80.70)16.70 (14.93, 18.63)6.36 (5.25, 7.68)9.96 (8.56, 11.54)Age Below 2074.40 (69.97, 78.37)0.5116.04 (4.11, 8.78)0.1247.00 (4.91, 9.90)0.1142.66 (1.48, 4.73)0.372 20—2977.01 (75.63, 78.32)9.27 (8.39, 10.24)6.02 (5.30, 6.83)4.34 (3.74, 5.04) 30—3976.25 (74.52, 77.89)9.48 (8.39, 10.71)5.72 (4.87, 6.72)4.50 (3.74, 5.39) 40 and above74.95 (71.01, 78.52)8.22 (6.13, 10.94)8.41 (6.30, 11.16)4.70 (3.17, 6.91)Wealth index Poor75.43 (74.08, 76.72)0.0096.43 (5.72, 7.22) < 0.0017.07 (6.32, 7.89)0.0012.93 (2.46, 3.50) < 0.001 Middle79.78 (77.36, 81.99)7.00 (5.67, 8.62)4.41 (3.37, 5.75)2.59 (1.82, 3.68) Rich76.61 (74.64, 78.47)16.14 (14.55, 17.88)5.22 (4.30, 6.32)8.42 (7.24, 9.76)Residence Urban76.91 (74.58, 79.08)0.07615.84 (13.99, 17.89) < 0.0016.66 (5.45, 8.12)0.0018.22 (6.87, 9.80) < 0.001 Rural76.78 (75.61, 77.92)7.56 (6.87, 8.32)6.45 (5.81, 7.16)3.77 (3.28, 4.33) Refugee72.87 (69.33, 76.13)7.01 (5.29, 9.24)2.74 (1.74, 4.31)0.61 (0.23, 1.61)Religion Anglican78.40 (76.64, 80.07)0.0298.36 (7.28, 9.59) < 0.0015.51 (4.64, 6.54)0.0034.70 (3.89, 5.66)0.017 Catholic75.21 (73.62, 76.73)8.24 (7.30, 9.29)7.36 (6.48, 8.36)3.51 (2.91, 4.24) Muslim77.73 (74.53, 80.63)13.31 (11.00, 16.00)5.46 (4.02, 7.39)4.34 (3.07, 6.11) Other75.16 (72.68, 77.49)9.95 (8.40, 11.75)4.77 (3.72, 6.11)5.58 (4.43, 7.01)Region Central77.10 (74.53, 79.49) < 0.00117.10 (15.00, 19.44) < 0.0013.35 (2.44, 4.59) < 0.0016.06 (4.80, 7.63) < 0.001 East82.46 (80.52, 84.26)8.80 (7.50, 10.29)7.92 (6.69, 9.35)7.29 (6.11, 8.68) North73.90 (72.13, 75.59)6.95 (6.01, 8.02)10.02 (8.90, 11.27)3.92 (3.22, 4.76) West74.40 (72.41, 76.29)7.48 (6.39, 8.73)1.38 (0.95, 2.01)1.43 (0.99, 2.07)CI confidence interval, p chi-square test p-value Table 2Proportions (95% CI) of women knowledgeable about malaria prevention methods by demographic characteristicsCharacteristicBed net useITN useTaking preventive medicineIRSProportion (95% CI)pProportion (95% CI)pProportion (95% CI)pProportion (95% CI)pOverall76.45 (75.45, 77.42)-9.08 (8.44, 9.77)-6.15 (5.61, 6.73)-4.32 (3.87, 4.82)-Education level None67.25 (64.82, 69.58) < 0.0016.22 (5.10, 7.56) < 0.0017.95 (6.69, 9.44)0.0021.14 (0.71, 1.82) < 0.001 Primary78.95 (77.67, 80.17)7.22 (6.47, 8.06)5.40 (4.75, 6.14)3.34 (2.83, 3.94) Secondary or higher78.74 (76.64, 80.70)16.70 (14.93, 18.63)6.36 (5.25, 7.68)9.96 (8.56, 11.54)Age Below 2074.40 (69.97, 78.37)0.5116.04 (4.11, 8.78)0.1247.00 (4.91, 9.90)0.1142.66 (1.48, 4.73)0.372 20—2977.01 (75.63, 78.32)9.27 (8.39, 10.24)6.02 (5.30, 6.83)4.34 (3.74, 5.04) 30—3976.25 (74.52, 77.89)9.48 (8.39, 10.71)5.72 (4.87, 6.72)4.50 (3.74, 5.39) 40 and above74.95 (71.01, 78.52)8.22 (6.13, 10.94)8.41 (6.30, 11.16)4.70 (3.17, 6.91)Wealth index Poor75.43 (74.08, 76.72)0.0096.43 (5.72, 7.22) < 0.0017.07 (6.32, 7.89)0.0012.93 (2.46, 3.50) < 0.001 Middle79.78 (77.36, 81.99)7.00 (5.67, 8.62)4.41 (3.37, 5.75)2.59 (1.82, 3.68) Rich76.61 (74.64, 78.47)16.14 (14.55, 17.88)5.22 (4.30, 6.32)8.42 (7.24, 9.76)Residence Urban76.91 (74.58, 79.08)0.07615.84 (13.99, 17.89) < 0.0016.66 (5.45, 8.12)0.0018.22 (6.87, 9.80) < 0.001 Rural76.78 (75.61, 77.92)7.56 (6.87, 8.32)6.45 (5.81, 7.16)3.77 (3.28, 4.33) Refugee72.87 (69.33, 76.13)7.01 (5.29, 9.24)2.74 (1.74, 4.31)0.61 (0.23, 1.61)Religion Anglican78.40 (76.64, 80.07)0.0298.36 (7.28, 9.59) < 0.0015.51 (4.64, 6.54)0.0034.70 (3.89, 5.66)0.017 Catholic75.21 (73.62, 76.73)8.24 (7.30, 9.29)7.36 (6.48, 8.36)3.51 (2.91, 4.24) Muslim77.73 (74.53, 80.63)13.31 (11.00, 16.00)5.46 (4.02, 7.39)4.34 (3.07, 6.11) Other75.16 (72.68, 77.49)9.95 (8.40, 11.75)4.77 (3.72, 6.11)5.58 (4.43, 7.01)Region Central77.10 (74.53, 79.49) < 0.00117.10 (15.00, 19.44) < 0.0013.35 (2.44, 4.59) < 0.0016.06 (4.80, 7.63) < 0.001 East82.46 (80.52, 84.26)8.80 (7.50, 10.29)7.92 (6.69, 9.35)7.29 (6.11, 8.68) North73.90 (72.13, 75.59)6.95 (6.01, 8.02)10.02 (8.90, 11.27)3.92 (3.22, 4.76) West74.40 (72.41, 76.29)7.48 (6.39, 8.73)1.38 (0.95, 2.01)1.43 (0.99, 2.07)CI confidence interval, p chi-square test p-value Proportions (95% CI) of women knowledgeable about malaria prevention methods by demographic characteristics CI confidence interval, p chi-square test p-value Effect of cumulative exposure to media channels for malaria messages on knowledge of malaria prevention Figure 1 illustrates a clear statistical difference between the cumulative number of malaria message media channels and women’s knowledge of ITNs, preventive medicine, and IRS. Mean channel exposure was significantly higher among women knowledgeable of these methods relative to their counterparts who were not (p < 0.05, t-test). Probability estimates of malaria knowledge (Fig. 2) showed a relationship between cumulative exposure to media channels and the increased knowledge of ITNs, preventive medicine, and IRS. No statistically significant effect of cumulative exposure was observed for knowledge of bed nets.Fig. 1Mean (95% CI) of the cumulative exposure to media channels for malaria messages by knowledge of malaria prevention method (Bed net, ITN, preventive medicine, and IRS)Fig. 2Probability of knowledge of malaria prevention methods (Bed net, ITN, preventive medicine, and IRS) by cumulative exposure to media channels for malaria messages Mean (95% CI) of the cumulative exposure to media channels for malaria messages by knowledge of malaria prevention method (Bed net, ITN, preventive medicine, and IRS) Probability of knowledge of malaria prevention methods (Bed net, ITN, preventive medicine, and IRS) by cumulative exposure to media channels for malaria messages Individual and household factors influencing cumulative exposure to media channels for malaria messages Multivariable mixed-effects negative binomial regression models (Table 3) identified both individual-level and contextual correlates of cumulative exposure. Education emerged as the strongest predictor: compared to women with no formal education, those with primary and secondary or higher education exhibited a 34% (IRR = 1.34; 95% CI 1.18–1.51) and 86% (IRR = 1.86; 95% CI 1.59–2.17) greater incidence rate of exposure to additional media channels, respectively.Table 3Unadjusted and adjusted incident rate ratios (IRRs) for mixed effects negative binomial models assessing demographic factors influencing the cumulative exposure to media channels for malaria messagesCharacteristicBivariateMultivariableIRR (95% CI)pIRR (95% CI)pEducation level No educationRefRef Primary1.27 (1.12, 1.43) < 0.0011.34 (1.18, 1.51) < 0.001 Secondary or higher1.88 (1.63, 2.17) < 0.0011.86 (1.59, 2.17) < 0.001Age Below 20RefRef 20—291.21 (1.01, 1.44)0.0401.17 (0.98, 1.40)0.082 30—391.33 (1.11, 1.60)0.0021.36 (1.14, 1.64)0.001 40 and above1.58 (1.27, 1.97) < 0.0011.69 (1.35, 2.11) < 0.001Wealth index PoorRefRef Middle1.20 (1.05, 1.36)0.0061.14 (1.00, 1.30)0.042 Rich1.61 (1.42, 1.82) < 0.0011.37 (1.19, 1.56) < 0.001Residence UrbanRefRef Rural0.67 (0.55, 0.82) < 0.0010.77 (0.63, 0.95)0.016 Refugee0.22 (0.15, 0.32) < 0.0010.28 (0.19, 0.42) < 0.001Religion AnglicanRefRef Catholic1.01 (0.91, 1.12)0.8761.02 (0.92, 1.13)0.772 Muslim0.96 (0.82, 1.13)0.6460.96 (0.82, 1.12)0.595 Other0.91 (0.80, 1.04)0.1650.92 (0.81, 1.05)0.235Region CentralRefRef East1.18 (0.90, 1.55)0.2191.66 (1.28, 2.16) < 0.001 North0.93 (0.73, 1.20)0.5971.86 (1.43, 2.42) < 0.001 West0.60 (0.46, 0.77) < 0.0010.88 (0.68, 1.13)0.318 Table 3Unadjusted and adjusted incident rate ratios (IRRs) for mixed effects negative binomial models assessing demographic factors influencing the cumulative exposure to media channels for malaria messagesCharacteristicBivariateMultivariableIRR (95% CI)pIRR (95% CI)pEducation level No educationRefRef Primary1.27 (1.12, 1.43) < 0.0011.34 (1.18, 1.51) < 0.001 Secondary or higher1.88 (1.63, 2.17) < 0.0011.86 (1.59, 2.17) < 0.001Age Below 20RefRef 20—291.21 (1.01, 1.44)0.0401.17 (0.98, 1.40)0.082 30—391.33 (1.11, 1.60)0.0021.36 (1.14, 1.64)0.001 40 and above1.58 (1.27, 1.97) < 0.0011.69 (1.35, 2.11) < 0.001Wealth index PoorRefRef Middle1.20 (1.05, 1.36)0.0061.14 (1.00, 1.30)0.042 Rich1.61 (1.42, 1.82) < 0.0011.37 (1.19, 1.56) < 0.001Residence UrbanRefRef Rural0.67 (0.55, 0.82) < 0.0010.77 (0.63, 0.95)0.016 Refugee0.22 (0.15, 0.32) < 0.0010.28 (0.19, 0.42) < 0.001Religion AnglicanRefRef Catholic1.01 (0.91, 1.12)0.8761.02 (0.92, 1.13)0.772 Muslim0.96 (0.82, 1.13)0.6460.96 (0.82, 1.12)0.595 Other0.91 (0.80, 1.04)0.1650.92 (0.81, 1.05)0.235Region CentralRefRef East1.18 (0.90, 1.55)0.2191.66 (1.28, 2.16) < 0.001 North0.93 (0.73, 1.20)0.5971.86 (1.43, 2.42) < 0.001 West0.60 (0.46, 0.77) < 0.0010.88 (0.68, 1.13)0.318 Unadjusted and adjusted incident rate ratios (IRRs) for mixed effects negative binomial models assessing demographic factors influencing the cumulative exposure to media channels for malaria messages Age was also positively associated with exposure to malaria messaging in multiple media channels. Women aged 30–39 years (IRR = 1.36; 95% CI 1.14–1.64) and those aged 40 years and above (IRR = 1.69; 95% CI 1.35–2.11) had significantly higher cumulative exposure relative to women younger than 20 years. Household wealth showed the effects of cumulative exposure to media channels, with women in the “rich” category exhibiting a 37% higher incidence of exposure (IRR = 1.37; 95% CI 1.19–1.56) compared to their “poor” counterparts. By contrast, women residing in rural areas (IRR = 0.77; 95% CI 0.63–0.95) and refugee settlements (IRR = 0.28; 95% CI 0.19–0.42) were significantly less likely to report exposure to multiple media channels compared to their urban counterparts. There was regional variation, with higher exposure observed in the eastern (IRR = 1.66; 95% CI 1.28–2.16) and northern regions (IRR = 1.86; 95% CI 1.43–2.42), while the western region did not differ significantly from the central region after covariate adjustment. Discussion The present study demonstrates that cumulative exposure to multiple media channels is significantly associated with enhanced knowledge of selected malaria prevention methods, particularly insecticide-treated nets (ITNs), preventive medicine, and indoor residual spraying (IRS). These findings reaffirm the central role of health communication in malaria control, highlighting that not only the presence but also the breadth of exposure across diverse media platforms matters for effective knowledge dissemination. The relationship observed in this study underscores the value of reinforcing malaria prevention messages through multiple channels to overcome knowledge gaps. The findings of this study are consistent with previous research showing that knowledge of malaria prevention methods is strongly associated with exposure to multiple media channels. Previous studies have demonstrated that awareness of key prevention methods, such as the use of insecticide-treated nets (ITNs), is positively linked to exposure to malaria messages [19, 20]. Health information delivered through diverse channels, including radio, television, poster/billboard, community event, interpersonal communication, flyers, social mobilization, social media, and community health workers, has consistently been shown to enhance knowledge of malaria prevention practices [10, 21]. In this study, exposure to multiple media channels was found to increase women’s knowledge of insecticide-treated nets (ITNs), preventive medicine, and indoor residual spraying (IRS). These findings are consistent with previous studies that have shown the positive effect of cumulative exposure to diverse media channels on both knowledge of malaria prevention measures and the subsequent adoption of these practices [22, 23]. Moreover, greater knowledge of different prevention methods has been shown to translate into a higher likelihood of their use, underscoring the role of information as a catalyst for behavior change [24, 25]. Taken together, this evidence suggests that cumulative exposure to malaria messages across multiple channels strengthens knowledge of prevention methods and supports their uptake in endemic settings [22, 23, 25]. Different factors at the individual and household levels influence exposure to malaria message media channels. In this study, education emerged as the strongest determinant of cumulative exposure, with women who had attained at least primary education being significantly more likely to report exposure to multiple channels compared to those with no formal education. This finding is consistent with evidence from studies in Ghana and Nigeria, which also demonstrated a positive association between higher educational attainment and exposure to malaria-related health messages across diverse media platforms [12, 26]. Age was another significant predictor, with older women more likely to have cumulative exposure compared to their younger counterparts. Similarly, research from Ghana has shown that knowledge of malaria prevention methods increases with age, partly due to accumulated exposure to health messages over time [22]. Several disparities were found to limit exposure to multiple media channels for malaria messages in this study. Consistent with evidence from previous research, women from wealthier households had greater cumulative exposure, underscoring the role of socioeconomic status in shaping access to health information [23, 27]. Place of residence also had a significant influence: women living in rural areas and refugee settlements were significantly less likely to report exposure to multiple channels compared to their urban counterparts. These findings highlight the need for malaria health communication strategies that are context-specific and equity-driven, employing tailored approaches to reach populations in resource-constrained and malaria-endemic settings [28]. Conclusively, the observed disparities by education, wealth, and place of residence in this study emphasize the persistent structural inequities in access to health information, reflecting broader social determinants of health that shape the diffusion of preventive knowledge in malaria-endemic settings. It is important to note that the absolute effect of each malaria messaging channel could not be determined; hence, the cumulative effect was based on the assumption that exposure to each media channel carried the same effect. This was beyond the researchers’ control since the surveys that collected these secondary data did not assess this. However, the findings still provide evidence of the reinforcing effect of the cumulative exposure to these media channels on malaria prevention. Additionally, since these data were based on a cross-sectional design, causal effects could not be established. Conclusion This study demonstrates that cumulative exposure to multiple media channels significantly enhances women’s knowledge of key malaria prevention methods, particularly insecticide-treated nets, preventive medicine, and indoor residual spraying. Education and age emerged as the strongest predictors of exposure, while wealth, residence, and regional disparities highlighted persistent inequities in access to malaria health information. These findings emphasize the importance of designing integrated, context-specific communication strategies that leverage both mass media and community-based platforms to ensure equitable reach. Strengthening multi-channel health communication remains critical for improving knowledge and promoting sustained adoption of malaria prevention practices in endemic settings. However, this study utilized secondary data from UMIS. Thus, the study used only existing variables in the dataset to investigate those associated with cumulative exposure to multiple media channels. Other variables that could influence exposure to media, for example, cultural norms and values, attitudes and practices, knowledge of utilization of media channels, among others, are not available in the dataset, yet they may affect exposure to media channels.