Dietary aflatoxin exposure of lactating mothers of children 0–6 months in Makueni County, Kenya

Abstract The southeastern region of Kenya is prone to aflatoxin outbreaks, yet maternal and infant aflatoxin intake levels remain unclear. We determined dietary aflatoxin exposure of 170 lactating mothers breastfeeding children aged 6 months and below in a descriptive cross‐sectional study involving aflatoxin analysis of maize‐based cooked food samples (n = 48). Their socioeconomic characteristics, food consumption patterns and postharvest handling of maize were determined. Aflatoxins were determined using high‐performance liquid chromatography and enzyme‐linked immunosorbent assay. Statistical analysis was conducted using Statistical Package Software for Social Sciences (SPSS version 27) and Palisade's @Risk software. About 46% of the mothers were from low‐income households, and 48.2% had not attained the basic level of education. A generally low dietary diversity was reported among 54.1% of lactating mothers. Food consumption pattern was skewed towards starchy staples. Approximately 50% never treated their maize, and at least 20% stored their maize in containers that promote aflatoxin contamination. Aflatoxin was detected in 85.4% of food samples. The mean of total aflatoxin was 97.8 μg/kg (standard deviation [SD], 57.7), while aflatoxin B1 was 9.0 μg/kg (SD, 7.7). The mean dietary intake of total aflatoxin and aflatoxin B1 was 7.6 μg/kg/b.w.t/day (SD, 7.5) and 0.6 (SD, 0.6), respectively. Dietary aflatoxin exposure of lactating mothers was high (margin of exposure < 10,000). Sociodemographic characteristics, food consumption patterns and postharvest handling of maize variably influenced dietary aflatoxin exposure of the mothers. The high prevalence and presence of aflatoxin in foods of lactating mothers are a public health concern and calls for the need to devise easy‐to‐use household food safety and monitoring measures in the study area.


| INTRODUCTION
Aflatoxins are secondary metabolites of fungal origin. They are released as spores that can withstand a range of extreme environmental conditions (Kumar et al., 2021). Their occurrence around the globe depends on geographic, climatic, agronomic and agricultural factors (Mahato et al., 2019). They can enter foods pre or postharvest (A. Kumar et al., 2021), and subsequently be consumed by humans and animals. Due to their harmful effects on animal and human health, they are intensively studied (Akbar et al., 2019).
The first incidence of aflatoxins' potency was reported in 1960 upon the death of 10,000 Turkeys and ducklings in the United Kingdom (Bhat et al., 2010;Blount, 1961). Since then, several incidences of aflatoxin contamination have been reported worldwide (Pickova et al., 2021). In Kenya, the first fatal case was reported in 1981 in the southeastern region of the country (Machakos, Kitui, Mwingi and Makueni former districts) (Omara et al., 2021). Since then, several fatalities and frequent episodes of aflatoxin contamination have been reported in the country (Mutegi et al., 2018;Omara et al., 2021). To date, alerts of aflatoxin-contaminated maize still feature in Kenya's news. In Nying'uro's (2020) study, incidences of aflatoxin occurrence in hotspot areas in Kenya are expected to increase due to climate changes.
Among aflatoxin studies examined by Omara et al. (2021), maize and other foods, mainly cereals, animal milk and animal feeds, were given more focus in Kenya. Additionally, it was observed that samples collected from the southeastern region of Kenya were associated with higher levels of aflatoxin compared to samples from other regions. However, among the few human studies, serum and urine were given more focus, whereas breast milk was only analysed by Kang'ethe et al. (2017) and Maxwell et al. (1989). Similarly, the studies that estimated exposure of lactating mothers to dietary aflatoxins were also lacking, except for a study conducted by Obade et al. (2021) among pregnant mothers in Kisumu County, the western region of Kenya. From this, it is clear that exposure of lactating mothers was not given more focus in Kenya, yet it could be one of the issues that require public health attention in the study region. This is because studies, including that of Kumar et al. (2016) among others, have associated high levels of aflatoxin exposure with acute aflatoxicosis characterised by haemorrhage, oedema and acute liver damage, while prolonged intakes, on the other hand, in small doses, have been shown to cause chronic aflatoxicosis, cancer, birth abnormalities in the foetus, malnutrition and immune suppression.
That notwithstanding, it has also been shown that higher levels of aflatoxins in mothers' diets could lead to higher levels of aflatoxin M1 in breast milk (Mehta et al., 2021), and, consequently, have a higher negative health effect on breastfeeding children compared to adults (A. Kumar et al., 2021). Additionally, this could also have an impact on the quality of breast milk that children aged 6 months and below solely depend on. According to Boquien (2018), breast milk is important during the first 6 months of life, is considered safe and acts as the primary source of nutrition for the young ones before they can eat other foods. Compromised breast milk safety could thus be one of the factors that could cause serious challenges in improving rates of exclusive breastfeeding in Kenya for children aged 6 months and below. However, at the time of conducting this study, information regarding the aflatoxin exposure of lactating mothers breastfeeding children of this age category remains scanty.
To reduce the negative impact associated with aflatoxin, it is important to determine the magnitude and reduce the risk of exposure. However, the government's effort to ensure the implementation of food safety and control measures for aflatoxins in Kenya still faces several challenges, especially among small and medium enterprises dealing with maize (Joutsjoki & Korhonen, 2021).
These challenges could still pose a threat to ensuring that lactating mothers are not exposed to aflatoxins through their diet. As a result, this study sought to determine dietary aflatoxin exposure of lactating mothers of children aged 6 months and below in Makueni County, Kenya. However, as aflatoxin analysis in breast milk was included in the larger part of this study, dietary aflatoxin exposure of mothers was also determined based on breastfeeding status.

| Sampling procedure
This study was part of a larger descriptive cross-sectional study with an analytical component conducted among lactating mother-child dyads in the Kikumbulyu location, Kibwezi-West Constituency, Makueni County, Kenya. A multistage sampling procedure was used to select the targeted households. The sample size of lactating mothers in the study was determined using Fisher et al.'s (1991) formula (n = Z 2 pq/d 2 ) where n is the desired minimum sample size; Z

Key messages
• A high prevalence of aflatoxin (84.4%) and high dietary aflatoxin exposure were reported among lactating mothers in the study.
• The mean concentration of 90% of maize-based cooked food samples exceeded the 2 and 10 µg/kg set limits for aflatoxin B1 and total aflatoxin, respectively.
• Household size, education, socioeconomic status and limited food consumption patterns were potential determinants of dietary aflatoxin exposure in lactating mothers. Within dietary diversity, the type of food constituting a diverse diet played an important role.
• Postharvest handling of maize could be one of the factors contributing to dietary aflatoxin exposure in the study area.
• The reported results call for the need to devise easy-touse household food safety and monitoring measures in the study area.
is the standard normal deviation set as 1.96 corresponding to 95% confidence interval (CI); p is the prevalence of aflatoxins maize samples above 10 µg/kg estimated at 87.3% in Makueni (IFPRI, 2010); q = 1 − p (proportion of maize sampled without aflatoxins), that is, 0.127; and d is the degree of accuracy set as 5%, that is, 0.05.
A sample size of 170 lactating mothers was arrived at as shown in the calculation: This present study targeted cooked maize-based food samples that were prepared and consumed by lactating mothers but remained or left for subsequent consumption at the time of data collection. To meet the calculated sample size (n = 40), at least five cooked maizebased food samples were purposively targeted per each of the eight wards in the study area. Any maize-based food sample was collected per household as long as it was consumed by the targeted mother.
However, the collection of food samples was based on chance (i.e., if maize-based food remained after a meal, and availability and the willingness of a lactating mother to give out the food samples during the EPI random walk and administration of study questionnaires).
Maize-based foods were purposively sampled in the study due to their high susceptibility to aflatoxin contamination.

| Ethical consideration
Ethical issues at all stages were considered in the study by obtaining ethical clearance (P454/08/2013) from Kenyatta National Hospital/ where Q is the estimated quantities of food consumed per sitting (g), F1 is the frequency of consumption within a typical day, F2 is the frequency of consumption per week, and 7 is the reference period of consumption frequency. Foods that were rarely consumed, for instance, once a month, were left out. It was also assumed that quantities of food were equally spread within a week in this calculation. Questions regarding the main source of maize, postharvest handling and storage of maize were also collected during the survey.
2.3.3 | Aflatoxin-prone foods weekly consumption score of each lactating mother The weekly consumption frequency score mainly for 18 selected foods that are likely to be contaminated by aflatoxins was determined (Supporting Information: Appendix 1). Foods that were consumed daily were given a weight of seven while those that were consumed twice, thrice or four times in that order until six times per week were given weights of two, three, four, five and six, respectively. A weight of zero was given to foods consumed per 2 weeks, per month and never as shown in Supporting Information: Appendix 1. The weights were multiplied by their respective consumption frequencies reported within a day to generate a total weekly consumption score as shown in the formula: Total weekly consumption score for aflatoxin − prone foods = Assigned food consumption weight × consumption frequency within a day.
The weighted aflatoxin consumption score for each lactating mother was subsequently derived by summing up the total weekly consumption score of all aflatoxin-categorised foods reported by the mothers and dividing the summed total consumed with a denominator of 504. The denominator was arrived at by multiplying the expected maximum food score per week (7) of each food by the expected maximum frequency consumption within a day (4) by the total number of foods (18) listed as foods likely to be contaminated by aflatoxins in this present study and expressing the result to a percentage shown as follows: Weighted aflatoxin consumption score per lactating mother = (∑[score food 1 + score food 2 +…+score food 18] × 100%)/504. Lactating mothers were further categorised into quartiles according to their percentage scores.

| Dietary diversity
A guideline for measuring dietary diversity (FAO, 2011) was used to generate 24-h dietary diversity scores for lactating mothers.
The score of 13 food groups (cereals, roots and tubers, vitamin A-rich vegetables and tubers, dark green leafy vegetables, other vegetables, vitamin A-rich fruits, other fruits, organ meat, flesh meats, eggs, fish and seafood, legumes, nuts and seeds, milk and milk products) were aggregated into 9 food groups where cereals and white tubers and roots were combined into starchy staples, other vitamin A-rich fruits and vegetables formed one group, other fruits, and other vegetables formed another group, and meat and fish formed a single group to generate women's dietary diversity score 3 (WDDS). Lactating mothers who had WDDS ranging between 1 and 3 were categorised as mothers with low dietary diversity, while the ones with scores between 4 and 6 and 7 and 9 were categorised as mothers with medium and high dietary diversity, respectively.

| Food samples
Solid food samples were collected using the quartering method (Campos-M & Campos-C, 2017). A representative sample was drawn from the top, middle and bottom of a plate, bowl or cup. Semiliquid foods were stirred to mix and a representative sample was scooped from the middle. The samples were transferred into a weighed cup until a 60 g sample weight was attained (50 g for quantification and 10 g for detection). The samples were stored in an ice-filled cooler box at temperatures of 4°C for 3 days (restocking of ice was done daily at the end of every data collection day) before being transferred to a deep freezer at −18°C. Food samples collected were solid maize meal (ugali, n = 18), semiliquid maize porridge (n = 6), maize-sorghum porridge (n = 9) and a mixture of boiled maize and beans (githeri, n = 9) and a mixture of dehulled maize and beans (muthokoi, n = 6). The varying proportion of food samples was a result of picking foods that remained after the household had had their specific meal at the time of data collection.

| Detection of positive aflatoxin in cooked maize-based food samples
Aflatoxin was detected as trifluoracetic derivatives using highperformance liquid chromatography (HPLC) and fluorescence detector (Nexera X2 Model; Shimadzu). For this process, 5 g of the samples were ground to fineness. Extraction was done using 25 mL of 70% methanol. The cleaning-up procedure was done 3 Dietary diversity score food groups used in this study were as per FAO (2011)  2.4.2 | Determination of total aflatoxin and aflatoxin B1 in positive food samples Positive samples were quantified for total aflatoxin and aflatoxin B1 using Ridascreen ELISA competitive enzyme immunoassay (r-Biopharm) with slight modification. Samples collected (50 g) were ground and mixed with 250 mL methanol-water mixture (70%:30%, v/v) and homogenised for 3 min for extraction. The resulting solution was filtered using Whatman filter paper number 1, and 50 µL was used for each standard and sample per well. Provided conjugate and antibody (50 µL each, respectively) were added, and incubation was done for 30 min at room temperature (25°C).
A wash buffer (phosphate buffer with tween) of 250 µL was used.
Chromogen (100 µL) was added as substrate and incubated again for 15 min at room temperature. Acid stop solution was added and the reading was done within 15 min. The recovery rate was set at 85% for total aflatoxin, and 93% for aflatoxin B1, while the absorbance reading was determined at 450 nm. However, the lower detection limit for total aflatoxin and aflatoxin B1 was set at 1.75 and 0.5 µg/kg, respectively.

| Determination of dietary aflatoxin intake of lactating mothers
Dietary aflatoxin intake was determined as shown in the formula: Aflatoxin intake (μg/kg/kgb. w. t/day) = Aflatoxin concentration(μg/kg) × Estimate quantities of food consumed (g)/day Body weight of lactating mother (b. w. t)(kg).
The contribution of each analysed food on the cumulative total aflatoxin and aflatoxin B1 intake in the study was determined based on regression coefficients derived by running the @Risk simulation model using the mean and range of aflatoxin intake for each food, risk β general distribution and 10,000 iterations.
Aflatoxin intake(μg/kg/kgb. w. t/day) = Aflatoxin concentration (μg/kg) × Estimate quantities of food consumed (g)/day Body weight of lactating mother (b. w. t)(kg). However, when the MOE was below 10,000, a value extreme from <10,000 indicated a higher risk of dietary aflatoxin exposure and a closer value to <10,000 indicated a lesser risk of dietary aflatoxin exposure in the study.

| Sociodemographic and economic characteristics of the study subjects
The mean household size of lactating mothers was 6.2 (SD, 1.3), while the mean number of children was 3.0 (SD, 1.7). About 90% of the lactating mothers interviewed (n = 170) were in the age category of 20 to 39 years with a mean age of 29.5 (SD,5.9) years (

| Consumption frequency of foods among lactating mothers
The consumption frequency of foods likely to be contaminated with aflatoxins in the study area is shown in Supporting Information: Consumption quantities of maize-sorghum porridge, 'githeri', 'muthokoi', cassava, finger millet, rice and groundnuts were less than 100 g/day. Consumption quantities of animal-based foods were less than 20 g/day (Table 2a).

| Dietary diversity of lactating mothers
The mean women's dietary diversity score in the study was 3.4 (SD, 1.5; range 1-6
Results also showed that the MOE of lactating mothers based on BMDL (0.41 μg/kg/b.w.t/day) against mean and 95th percentile levels of total aflatoxin and aflatoxin B1 dietary intake as shown in Supporting Information: Table S4 were lower than 10,000 cut-off point adopted by EFSA (2005).

| Correlation of variables with dietary intake of total aflatoxin and aflatoxin B1 of lactating mothers in Kibwezi West
As shown in Table 5a, a significant negative correlation was reported between the education level of exclusively breastfeeding mothers and the intake of total aflatoxin in the study (ρ = −0.47, p = 0.040).
A negative significant correlation was reported between the dietary diversity of nonexclusively breastfeeding mothers with total aflatoxin intake (t b = −0.36, p = 0.03). Despite cleaning or sorting maize before storage, a positive significant correlation was reported with the intake of total aflatoxin among non-exclusively breastfeeding mothers (t b = 0.39, p = 0.037). No significant correlation was observed with the remaining variables.
As shown in Table 5b 3.6 | Predictors of aflatoxin exposure of lactating mothers of children 0-6 months Among predictor variables, as shown in Table 6, the level of education was found to significantly and negatively influence estimates of aflatoxin B1 intake among exclusively lactating mothers in the study (β = −0.56, p = 0.01). Women's dietary diversity scores were found to negatively influence estimates of aflatoxin B1 intake among nonexclusively lactating mothers (β = −0.43, p = 0.04). The regression coefficient using the @Risk simulation model showed maize ugali as the greatest contributor to cumulative intake of both total aflatoxins (b = 0.69) and aflatoxin B1 intake (b = 0.70). The least contributor to the cumulative intake of both total aflatoxin and aflatoxin B1, however, was reported for 'muthokoi' (b = 0.02) (Supporting Information: Figure S1).

| DISCUSSION
The observed high mean household size of lactating mothers compared to the national mean was expected as rural areas in Kenya are generally characterised by larger household sizes   According to weekly consumption frequency, 'ugali', maize porridge, 'githeri' and groundnut would put lactating mothers at a higher risk of aflatoxin exposure. This did not come as a surprise as these foods constitute staple foods mostly consumed in Kenya.
Other foods that could easily contribute to aflatoxin exposure include animal milk (mostly cow's milk) in the form of milk tea and 'muthokoi'.
However, their frequency of consumption was modest per week in the study. As fish, chicken, cassava, finger millet, eggs, plain sorghum flour and mixed flour porridge were rarely consumed per week in the study, they were considered to contribute less to dietary aflatoxin exposure among lactating mothers. Overall, the food frequency result of this study was similar to those of Kilonzo et al. (2014)  This study agrees with Koskei et al. (2020) and Malusha et al. (2016) that metal and plastic buckets, sacks and granaries are the commonly used maize storage containers in Makueni. Okoth and Ohingo (2004) further showed that storage in plastics, polythene bags, metal buckets and sacks leads to a moisture content of at least 13.6% in maize. In the environment of a hot and humid temperature, it is highly possible that storage practices could be a cause of the high prevalence of aflatoxin in Makueni. Just as the study of Nii et al. (2019), the storage of maize products for a long time in these containers could exacerbate the occurrence of aflatoxin in the area.
The findings of this study showed that most households in Makueni bought maize from the market, followed by their cultivation and T A B L E 2b Source, storage and processing of cereals by lactating mothers in Kibwezi West. other sources (gifts/donations/reliefs). These findings were similar to the ones reported by Daniel et al. (2011). However, varying results linking aflatoxin contamination and source have been reported in the study area. For instance, the study by Daniel et al. (2011) showed that home-produced maize was highly contaminated by aflatoxin compared to the ones bought from the market and those given out as relief food. On the other hand, Mwihia et al. (2008) showed that maize produced at home and those bought from the market were all highly contaminated with aflatoxin. This study, however, did not categorise the food samples into sources, and, thus, could not link the source of maize and aflatoxin occurrence in the study. Lastly, the proportion of lactating mothers applying methods that could reduce aflatoxin occurrence in maize was found to be lower than those reported by Koskei et al. (2020). It is thus clear that maize handling and storage practices among lactating mothers are still low to mitigate the occurrence of aflatoxin in the study area.  (2015). In this experiment, it was concluded that aflatoxin ratios within the food matrix will always vary between those reported for total aflatoxin and individual types of aflatoxin. However, the high proportion of food samples exceeding the KEBS limit is alarming and should be a cause of food safety concerns in the study area.
The mean dietary intake of total aflatoxin (7600 ng/kg b.w.t/day, i.e., converted from μg/kg b.w.t/day) among lactating mothers in this study was remarkably higher than those reported by Kilonzo et al.  (Liu & Wu, 2010), and surpassed the upper bound exposure levels of aflatoxin B1 intake of 3.25 ng/kg b.w.t/day estimated for adults from several studies (EFSA CONTAM Panel et al., 2020). The levels were equally higher than the 1.19 ng/kg b.w.t/day reported in Ghana (Kabak, 2021). The high intake levels reported in this study could be because this present study also factored in the consumption frequency of foods within a day. However, going by these results, it is evident that dietary intake of both total aflatoxin and aflatoxin B1 is higher among lactating mothers in the study area. The high intake was greatly attributed to the high consumption of maize 'ugali' followed by maize porridge, maize sorghum porridge, 'githeri' and Total aflatoxin intake (μg/kg b.w.t/day) All mothers EBF mothers NEBF mothers Food x Maize ugali a 14.4 (6.2), n = 14 3.6-23.9 16.7 (6.1), n = 8 3.6-23.9 11.2 (5.3), n = 6 4.7-17.6 Maize porridge 4.0 (6.2), n = 6 0.3-16.1 1.68, n = 1 1.68 4.5 (6.7), n = 5 0.   Leroy et al. (2015) but contrasts with the findings of Mehta et al. (2021). That notwithstanding, findings shared by Malusha et al. (2016) and Lesuuda et al. (2021) showed a negative correlation between knowledge, attitude and practices with aflatoxin contamination on cereals, and underscores education as one of the strategies that can be implemented among lactating mothers to reduce dietary aflatoxin exposure in this present study. Though the socioeconomic status model did not predict aflatoxin B1 intake of lactating mothers in the study, its positive association with an increase in aflatoxin B1 intake was not expected. This is because the results of Leroy et al. (2015), Nabwire et al. (2020) and a review by Omara et al. (2021), among others, had associated households with lower socioeconomic index with a higher probability of aflatoxin exposure within the same study area of this present study. Similarly, this study showed that almost 95% of lactating mothers in the study were below the upper wealth index.
Further investigation for a possible explanation was conducted.
Results of this study though nonsignificant, showed a positive association between socioeconomic status and an increase in the total number of meals consumed per day by lactating mothers and an increase in women's dietary diversity in the study. But a predictor model showed that women dietary diversity negatively influenced aflatoxin B1 intake among non-exclusively lactating mothers. This outcome which showed that low dietary diversity was a risk factor for aflatoxin intake, was also consistent with those reported by Leroy et al. (2015) and Nabwire et al. (2022) conducted in the same area of this study but inconsistent with those reported away from this study area by Mehta et al. (2021) in India. However, when focusing on foods that are only susceptible to aflatoxin contamination, the study by Andrews-Trevino et al. (2020) reported a positive correlation between socioeconomic status and consumption frequency of contaminated maize and groundnuts among Nepalese women. This study, therefore, suggests that the influence of socioeconomic status on aflatoxin levels in foods depends on the region of the study, the available type and range of food diversity and the prevalence of aflatoxin contamination in the area. When used on its own, it might not be a reliable pointer of aflatoxin intake in a study area. For instance, without basic knowledge of aflatoxin contamination and limited food choice, lactating mothers with higher socioeconomic status can still be susceptible to aflatoxin B1 exposure in the diet. Even though Andrews-Trevino et al. (2020) reported a negative association between age and exposure to aflatoxin B1 in the serum of pregnant mothers, this study, regardless of the breastfeeding status, did not find any direct influence of age of the lactating mothers on the total aflatoxin and aflatoxin B1 levels in the study area. This could probably be due to a range of sociodemographic and economic similarities drawn between exclusively and nonexclusively lactating mothers in the study.

| CONCLUSION
The high prevalence and presence of aflatoxin in foods of lactating mothers are a public health concern and calls for the need to devise easy-to-use household food safety and monitoring measures in the study area. However, health practitioners, policymakers and food safety experts should also scale awareness among lactating mothers, and small -and -medium-scale maize and T A B L E 6 Predictors of aflatoxin exposure among lactating mothers breastfeeding children 0-6 months in Kibwezi West.