Bungoma County, situated in western Kenya in the Lake Victoria Basin, encompasses an area of 3032.4 km² and is positioned between latitudes 00˚28'N and 10˚30'N, and longitudes 34˚20'E and 35˚5'E. The county has a population of around 1.67 million and borders Uganda to the northwest, Trans-Nzoia to the northeast, Kakamega to the east, and Busia to the west. It is made up of nine sub-counties: Kanduyi, Bumula, Sirisia, Kabuchai, Mt. Elgon, Webuye East, Webuye West, Tongaren, and Kimilili, see Figure 1 . ( Tsimbasi et al., 2024 ; Bungoma County, 2021 )

Its fertile volcanic soils supporting extensive agriculture may contain natural radionuclides like ²²⁶Ra, ²³²Th, and ⁴⁰K, which could be further concentrated with fertilizers. Bungoma’s population primarily depends on locally grown maize and beans, so it is important to evaluate the radiological risks that may be associated with consuming them. Agriculture is the county’s main economic activity, and these crops are essential for both subsistence and trade.

In each of the nine sub-counties, 200 g samples of beans and maize were chosen at random from a subset of households in a single season. To prevent cross-contamination, the samples were placed in tiny, sterile bags, identified with the GPS location, and brought to the lab for processing. To make sure the moisture content was eliminated, the samples were first sun-dried for two days. Nine representative samples were created by mixing 150 g of each of the three maize and bean samples from each sub-county. After the materials were ground into a powder, they were filtered through a 2 μm sieve to create a finely grained sample. The samples were subsequently dried in an oven at 50˚C to 60˚C until a stable weight was attained. Samples were considered dehumidified when the rate of mass variation went to zero ( Lopes et al., 2018 ). 200 g of representative samples were weighed using an electronic top pan balance, put into cylindrical plastic containers, sealed tightly, and labeled suitably. Table 1 lists the labels for the maize samples, which were M₁ through M₉. Similarly, bean samples were designated B₁ through B₉ as highlighted in Table 2 . To allow ²²⁶Ra and ²³²Th to achieve a secular equilibrium with their daughter radionuclides, the samples were stored for a month ( El-Gamal et al., 2019 ).

International Atomic Energy Agency (IAEA) standard reference materials RG-U, RG-K, and RG-Th were used to acquire the spectrum and estimate the activity concentration of specific radionuclides. The activity concentration of ²³²Th was determined using the ²⁰⁸Tl (2615 keV) gamma energy peak, while the activity concentration of ²²⁶Ra was estimated using the activity concentration of ²¹⁴Bi (1765 keV). Because the energy peaks are different from other peaks, they were selected to compensate for the gamma-ray detector’s low level of resolution. Its gamma energy peak at 1460 keV was used to determine the activity content of ⁴⁰K. Each sample’s data collection time was 28,800 seconds.

Determining the background radiation was a necessary step in establishing the minimum activity that could be reliably detected ( Sultana et al., 2019 ). This involved measuring counts from distilled water placed in a plastic container over the same time period as the sample measurements. The resulting background counts were deducted from the total counts to yield net counts, which were then utilized to calibrate the energy scale.

The decay rate of the radionuclide per unit mass is known as a specific activity. The precise activity linked to every radionuclide in a sample is given by Equation (1) ( Wabomba et al., 2022 ; Yarima et al., 2019 ).

$C=\frac{N\times 100}{P_{\text{γ}}\times \text{ε}\times m}$ (1)

where C is the activity of the sample in Bq/kg, ε is the counting efficiency of the gamma energy, P_γ is the absolute intensity of the gamma-ray and m, net mass of the sample (in kilograms).

The whole-body health risk resulting from food consumption is represented by the annual effective ingestion dose ( Bilgici Cengiz & Caglar, 2022 ). Equation (2) can be used to calculate a person’s yearly effective ingestion dose based on their dietary intake. According to the Kenya National Bureau of Statistics (2019) , maize consumption per capita was an average of 64.1 kg per year, 2014-2018.

AEID (μSv·y⁻¹) = C × A_I × F_DC (2)

where A_I is the annual intake of each food in kilograms per year. F_DC is the standard dosage conversion factor, which is equivalent to 0.28 μSv·Bq⁻¹ for ²²⁶Ra, 0.23 μSv·Bq⁻¹ for ²³²Th and 0.0062 μSv·Bq⁻¹ for ⁴⁰K. The coefficients account for age-related changes in intestinal absorption, body and organ masses, and urinary bladder excretion rates ( ICRP, 2012 ).

The gamma ray index (𝐼_𝛾) is used to calculate the degree of gamma radiation risk connected to naturally occurring radionuclides in particular materials, Equation (3) ( Geremew, 2023 ; Hameed & Fzaa, 2021 ).

$I_{\text{γ}}=\frac{A_{\text{Ra}}}{150}+\frac{A_{\text{Th}}}{100}+\frac{A_{\text{K}}}{1500}$ (3)

The lifetime cancer risk refers to the likelihood of developing cancer risk arising from foods taken with the daily diet. Lifetime risk assessment will be estimated by applying the annual ingestion effective dose due to the presence of ²²⁶Ra, ²³²Th, and ⁴⁰K radionuclides in the foods, from the age of first exposure until the actual life expectancy of 67.47 years, according to the Kenya National Bureau of Statistics (KNBS), Equation (4) ( Bilgici Cengiz & Caglar, 2022 ).

$\text{LTCR}=\text{AEID}\times L_{\text{span}}\times F_{RC}$ (4)

where L_span is the mean life span (average 67.47 years), F_RC is a risk conversion factor and the F_RC coefficient is used to convert internal doses to excess cancer risk are 5.02 × 10^–3, 7.38 × 10^–1 and 3.58 × 10^–1 μSv⁻¹ for ⁴⁰K, ²³²Th and ²²⁶Ra, respectively.

The average ²³²Th, ²²⁶Ra and ⁴⁰K activity in maize were 54.00 ± 21.15 with a range of (23 ± 1.1 - 75 ± 3.79) Bq/kg 20.90 ± 7.19 with a range of (11.50 - 30.10) Bq/kg, and 161.00 ± 76.84 with a range of (46 ± 2.34 - 279 ± 13.99) Bq/kg respectively as seen in Table 1 . The highest activity was recorded in ⁴⁰K in sample M₄, while the lowest was recorded from ²²⁶Ra in sample M₁.

The average activity level of radionuclide in the nine beans samples for ²²⁶Ra, ²³²Th, and ⁴⁰K were 18.40 ± 4.03 with a range of (11.5 ± 0.57 - 23.1 ± 1.15) Bq/kg, 43.00 ± 15.51 with a range of (23 ± 1.16- 67 ± 3.35) Bq/kg and 195.00 ± 132.48 with a range of (28 ± 1.4 - 498 ± 24.93) Bq/kg respectively as seen in Table 2 . The highest activity concentration in beans was recorded from ⁴⁰K in B₂, while the lowest was determined from ²²⁶Ra in B₉.

Soil composition plays a crucial impact as geological formations, including granitic and volcanic soils, tend to have higher quantities of radionuclides such as ²²⁶Ra, ²³²Th, and ⁴⁰K ( Marques et al., 2021 ). The usage of phosphate-based fertilizers in the study area, which frequently includes trace levels of radioactive materials, additionally contributes ( Kim & Cho, 2022 ). Environmental factors such as erosion, water runoff, and sediment transport redistribute radionuclides, resulting in uneven concentrations across agricultural fields ( UNSCEAR, 2000 ). Understanding these elements is critical for effectively managing radiological hazards in food and the environment.

The average AEID due to ingestion of maize was found to be 1.239 (0.585 - 1.621) mSv·y⁻¹, Table 1 , while that of beans was 0. 263 (0.156 - 0.36) mSv·y⁻¹, Table 2 ( Tsimbasi et al., 2024 ). Maize samples recorded higher AEID compared to bean samples, as seen in Figure 2 . The mean AEID values in maize were above the limit of 1 mSv·y⁻¹ for the general public, as recommended by the International Commission on Radiological Protection (ICRP). This shows that there is an increased health risk for the whole body of an individual due to the intake of maize with higher AEID.

The AEID values found in the current study for Bungoma are significantly greater than the 0.5 mSv/year noted in Brazil ( Lopes et al., 2018 ), but lower than the 6.13 mSv/year reported in Iran ( Changizi et al., 2013 ). Compared to the current study for Bungoma, studies from Nigeria show AEID values that are much lower, ranging from 1.9 μSv/year to 5.12 μSv/year ( Yarima et al., 2019 ). Tanzania’s maize AEID, on the other hand, was 8.64 × 10⁻² mSv/year ( Banzi et al., 2017 ), suggesting a closer exposure level. AEID values in the Bureti region of Kenya were 5.92 mSv/year ( Rotich, 2022 ) which was significantly higher than what was found in the present research. Although consumers may not be immediately in danger of any health problems due to the average concentration of all radioisotopes in maize and beans, there may be a long-term cumulative effect once the current dose is consumed ( Tsimbasi et al., 2024 ; Changizi et al., 2013 ; Muhammad et al., 2024 ). Increased radiation exposure to particular organs can affect human health by impairing immunity, which may increase the risk of radiation-related illnesses like cancer and ultimately raise rates of fatalities ( Tawalbeh et al., 2011 ).

The average gamma index, I_ϒr for maize and beans were 0.7 ± 0.03 (0.3 ± 0.01 - 1 ± 0.05) mSv and 0.6 ± 0.03 (0.3 ± 0.01 - 1 ± 0.05) mSv respectively. The highest value recorded in maize samples B₂ and M₉ exceeded the threshold value of 1 mSv, as seen in Figure 3 . The average values indicate that there are minimal risks posed to consumers due to the natural radionuclides present in maize and beans. The mean ELCR was 4.89 × 10⁻⁸, ranging from 2.22 × 10⁻⁸ to 6.53 × 10⁻⁸ in maize and 1.00 × 10⁻⁸, ranging from 5.9 × 10⁻⁹ to 1.42 × 10⁻⁸ in beans. Studies in Nigeria ( Yarima et al., 2019 ; Muhammad et al., 2024 ) revealed maize ELCR values ranging from 1.1 × 10⁻¹³ to 1.6 × 10⁻⁵, comparable to the results found in this study. In contrast, Onjefu et al. (2021) documented ELCR values well within internationally permissible limits in Namibia, less than those recorded in Bungoma County, further underscoring the need for regular monitoring.

Samples M₉ and B₂ recorded the highest values of Gamma index which can be related to higher specific activity concentration and AEID, Figure 2 and Figure 3 .

This study suggests that maize consumption presents a slightly greater radiological risk than beans, primarily due to raised AEID and ELCR values, which can be attributed to the higher absorption of ²³²Th from volcanic soils, differences in plant uptake mechanisms and variations in dietary intake patterns ( Tsimbasi et al., 2024 ). Although the overall radiation levels in both crops are within international safety limits, localized hotspots with higher radiation concentrations indicate potential long-term health risks, especially in maize-consuming populations. The presence of raised ²³²Th levels in both crops highlights the need for further soil analysis and monitoring to determine the underlying causes of contamination. Since long-term exposure to radionuclides may increase cancer risk, public health interventions should focus on regular monitoring, raising awareness about food safety, and implementing dietary diversification strategies to reduce excessive maize dependency. Conducting a geospatial survey of Bungoma’s farmland would help identify high-risk areas and guide mitigation efforts. Furthermore, research into farming practices, including the impact of fertilizers and soil composition, could help in reducing radionuclide uptake in food crops, ultimately ensuring long-term food safety and health security for the local population.