The use of DDT in Uganda dates back to the 1950s when it was used as the main pesticide against pests and vectors up to mid-1980’s (Nnamuyomba et al., 2014); although its use world-wide was stopped in early 1960s owing to international concern of its bioaccumulation effects and the subsequent resistance of vectors against the chemical (Wasswa et al., 2011). However, on behalf of the Ugandan government, the Ministry of Health by the approval of Uganda’s National Environment Management Authority (NEMA) (USAID, 2007) through the National Malaria Control Program (NMCP) (Acquaye and Chandonait, 2014), supported by president’s malaria initiative decided to re-introduce DDT use through IRS program in 2008 as a means to control malaria. In Uganda, malaria is highly endemic and it puts to approximately 90% of the population at risk and causes half of the inpatient pediatric deaths. To use DDT, the ministry argued that the insecticide is cheap, effective and its insecticidal properties are credited for saving over 100 million lives of people since 1944 to date globally mainly from malaria. Apac and Oyam districts which were among the most malaria-prone districts in the country were selected and sprayed by DDT in April 2008 to pilot the IRS programme (RTI, 2008).

The use of DDT in malaria control raises serious health and environmental concerns because the chemical and its metabolites can persist for years within the sprayed domestic environment (Van Dyk et al., 2010). Consequently, residents can be exposed to the residues through a number of pathways including inhaling indoor dust, outdoor soil and food intake (Tao et al., 2009) for instance, leafy vegetables around the sprayed homesteads might be contaminated with DDT residues by both root uptake from contaminated soils and spray-drift from application during IRS. Therefore, consumption of such contaminated vegetables makes the people susceptible to the numerous deleterious effects of the insecticide (Ssemugabo et  al., 2022)  because  ingestion of food has been recognized as the major pathway for human exposure to DDT (Tao et al., 2009).

After IRS programme in 2008 in Apac and Oyam districts, no comprehensive and justifiable study has been performed to determine the levels of DDT and its metabolites in the environment and agricultural produce in this area. Therefore, this study was aimed at assessing the levels of DDT residues in soils, leafy vegetables (food crops) around the sprayed homesteads and indoor soil dust from the sprayed houses of selected villages of Apac and Oyam districts of Northern Uganda.

Study area

The areas of study were the two districts (Apac (01°59′N 32°32′E) and Oyam (02°14′N 32°23′E)) in the Northern region of Uganda. The two districts are bordering each other, with Oyam being located to the Northwest of Apac (Figure 1). Inset is the location of the areas selected in the study in Uganda.

In Apac district, two villages: (Aboko parish) and Olamia A (Ongoceng parish); and Arocha swamp were selected for sampling, while in Oyam district, two villages: Abang and Arukolong (Pukica parish); and Tochi swamp were selected for sampling. The villages were randomly selected on the basis of ease to locate the sprayed homesteads which could be a representation of the 92.4 and 93.1% sprayed houses in Apac and Oyam districts, respectively (RTI, 2008), while the swamps were the ones at the lower ends of the villages sampled where runoffs are deposited over time.

Karuma Wildlife Reserve (KWR), which is part of Murchison Falls protected area (2° 14' 3"N, 32° 14' 47"E) was used as the control site. KWR is located in Kiryandongo district in Western Uganda. The KWR was chosen as the control site for the study of soils because the area was anticipated to be free from direct exposure of insecticides such as DDT since it was gazetted in 1964 (Behangana et al., 1997). Gwara village in Kiryandongo district was another control area for the study of indoor house dust, outdoor soils and leafy vegetables because no IRS in 2008 took place in that area. The homesteads were randomly selected for house indoor dust, outdoor soils and leafy vegetables sampling.

Sample collection

Soils (50 samples), indoor dust (25 samples) and the selected leafy vegetables (75 samples) around the selected sprayed homesteads were randomly and conveniently collected from selected villages in Apac and Oyam districts and from the control areas in June, 2020. Indoor dust samples were obtained by scratching the indoor walls using a chisel. The dust obtained was swept using a broom from the very house, collected and transferred into an aluminium foil then into well labelled sampling bags, and then placed into a clean dry bucket. The outdoor soil samples around the sprayed homesteads were obtained by judgemental sampling technique from the sampling points located in three different directions around the sprayed house at 2 to 5 m and/or 10 m away from the sprayed and control house at a depth of up to 10 cm using Auger handle, mixed thoroughly and a representative amount transferred into aluminium foil then into a well labelled sampling bag, and then put into a clean dry bucket. Soil samples from Arocha and Tochi swamps and KWR were obtained at approximately over 500 m away from the main roads in these areas,  20 m  apart  in  different  directions at a depth of up to 10cm using Auger handle. Samples of the selected leafy vegetables (Solenum lycopersicum L, Solenum melongena L, Hibiscus sabdariffa, Abelmoschus esculentus L and Gynandropis gynandra L) and or their fruits growing around the sprayed and control homesteads, were handpicked separately, wrapped into aluminium foil, transferred into a well labelled sampling bag, and placed into the clean dry bucket. All samples were then transported into the pesticide laboratory, Department of Chemistry, Makerere University, Uganda where they were kept in a freezer maintained at -18°C until extraction.

Chemicals and reagents used

The analytical standards (p,p’-DDT, o,p’-DDT, p,p’-DDE, o,p’-DDE, p,p’-DDD and o,p’-DDD) used, were obtained from Dr. Ehrenstorfer GmbH (Augsburg, Germany). Stock standard solutions obtained from the crystalline standards were diluted (0.5 to 5 μg/ml) using cyclohexane pesticide grade as solvent for analysis. All pesticide residue grade solvents of ethylacetate, acetone, cyclohexane, and n-hexane used were supplied by British drug houses (BDH, UK). The purity of the solvents for extraction was checked using gas chromatography-electron detector (GC-ECD) to ensure that there are no detectable trace levels of the pesticides of interest present in the solvents. Other chemicals such as sodium sulphate  and  florisil used was preheated at 130°C to remove moisture followed by deactivation of the later with 5% distilled water. Ammonium chloride and sodium chloride were also used.

Extraction of samples

Extraction and cleanup of soil and dust samples

Study samples were defrosted and airdried at room temperature before extraction. Soil and dust samples were extracted for DDT residues using the solid dispersion method (Åkerblom, 1995). 20 g of soil sample was weighed into a 250 ml Erlenmeyer flask (E-flask) and 14 ml ammonium chloride solution (0.2 M) added. The mixture was hand swirled, and then left to stand for 15 min. Cyclohexane/Acetone (1:1 v/v) solvent system (100 ml) was added and the resultant mixture vigorously shaken for a minute, then less vigorously at intervals of 10 min for 1 h. The mixture was then left to stand overnight, after which it was shaken intermittently for another 2 h and left to settle. Distilled water was then added cautiously until the organic phase filled the neck of the E-flask. A saturated sodium chloride solution (50 ml) was added and the mixture was shaken for 1 minute, allowed to separate into an upper organic phase and a lower aqueous phase. The upper organic phase was then collected, dried  using  20 g  sodium  sulphate in 15 min interval and decanted through a plug of glass wool into an evaporation flask. The extract was then subjected to a clean-up process using the florisil technique (Kegley and Laura, 1998). Briefly, a glass column (15 cm × 4 mm i.d) was plugged with glass wool rinsed with cyclohexane, packed with 10 g florisil followed by 4 g anhydrous sodium sulphate, then conditioned with 2 ml acetone/hexane (1:9 v/v), followed by hexane (5 ml). The soil extract was then added to the column, eluted with 10 ml acetone/hexane (1:9 v/v), followed by 10 ml hexane and the eluate was collected in a 100 ml evaporation flask. The eluate was then concentrated on a rotary evaporator at 40°C followed by a gentle stream of nitrogen gas to dryness and then reconstituted in cyclohexane (2 ml) for gas chromatograph (GC) analysis.

Extraction and cleanup of vegetable samples

A multi-residue method DFG S19 (Specht et al., 1995) for the determination of pesticides residues in vegetables and fruits was followed during the extraction of vegetables for DDT residues. 25 g of the chopped homogenized vegetable was mixed with acetone (50 ml), 50 ml ethyl acetate/cyclohexane (1:1 v/v) followed by 50 g of sodium sulphate and then blended for 5 min. The extract obtained filtered over sodium sulphate (10 g), 20 ml of saturated aqueous sodium chloride solution was added, mixed vigorously, hand shaken for 1 min, and allowed to settle. The upper organic layer was Pasteur pipetted into a flask containing sodium sulphate (15 g), the mixture was vigorously hand shaken for 30 s and then filtered over sodium sulphate through Whatman GF/C Glass Micro-fibre filters and rinsed with cyclohexane/ethyl acetate (1:1 v/v) solvent mixture. All the filtrates (supernatant) were transferred into a 100 ml evaporation flask, concentrated to about 4 ml on a rotary evaporator at 40°C and then divided into two portions (each of 2 ml), one for florisil clean up and another for acid treatment.

The clean-up of extracts from vegetable samples was by florisil (Åkerblom, 1995). Briefly, 10 g florisil packed in the column was conditioned using 5 ml acetone/hexane (1:9 v/v) followed by 5 ml hexane. The first 2 ml portion of the vegetable extract (obtained as described earlier) was dissolved in hexane (2 ml), eluted with 10 ml acetone/hexane (1:9 v/v). The eluate was then concentrated on a rotary evaporator at 40°C followed by a gentle stream of nitrogen gas to dryness and then reconstituted in cyclohexane (2 ml) for gas chromatograph (GC) analysis.

For acid treatment, five soil and five vegetable extracts were selected and treated with sulphuric acid as described by Åkerblom (1995) to access the stability of DDT and its metabolites in the samples analysed. DDT residues were detected in the acid treated portions.

GC chemical analysis of sample extracts

The GC (Varian 3800-CP) equipped with ECD and SPB-1 0.53 mm i.d. × 60 m non-polar capillary column was used for analysis of sample extracts and later interchanged with Wcot fused silica (30 m × 0.25 mm i.d) coated CP-Sil 19 CB semi-polar capillary column, for confirmation. The GC temperature program was as follows: initial temperature 90°C (held for 1 min), raised to 180°C at a rate of 30°C min^-1, and then raised to 260°C at a rate of 4°C min^-1 (held constant for 10 min), for both non-polar and semi-polar columns. The injection port and ECD detector temperatures were 230 and 300°C, respectively. 1 μl of the sample was injected, hydrogen gas and nitrogen gas at 1.2 and 30 ml min^-1 flow rates was used as the carrier and makeup gases, respectively. A Turbochrom (Perkin-Elmer Corporation, 1989-1995, USA) 4.0 chromatography work station was used for chromatographic data processing. The GC-ECD calibration linear plots were of peak area against the concentration of the standards. Identification of DDT residues was based on retention time as well as external standard method. Selected samples were analyzed by GC-MS for structural confirmation of the DDT residues.

Quality assurance and quality control

Recoveries, precision and accuracy were determined by addition of standard method at two spiking levels of 2 and 4 μg/ml of a mixture of standards (p,p’-DDT, o,p’-DDT,  p,p’-DDE, o,p’-DDE, p,p’-DDD and o,p’-DDD) into the soil (from KWR) and vegetable samples (S. melongena) obtained from control site, kept overnight, extracted as of the study samples and analysed in triplets (ICH, 1995; EPA, Method 8000C, 2003). The mean percentage recoveries for soil and vegetable samples with standard deviations are shown in Table 1.

From the percentage recoveries, the data was not corrected for use because the average recoveries obtained were within the acceptable limits of 70 to 120% (Akerblom, 1995) or 70 to 130% (USEPA, 2003). The close values of mean percent recoveries were also indices of good accuracy (IUPAC, 1994) while the low standard deviation values signified good precision (Ripp, 1996) of the analytical process used in the study.

To determine the lower limit of detection, LLD (sensitivity of GC-ECD), three soil samples from the control site were fortified at a lower concentration of 0.02 μg/ml of the mixture of standard DDT insecticide, and analysed in triplicate (ICH, 1995). LLD was estimated as 3.29 times standard deviation (Eaton et al., 1995) because for chromatographic techniques, the response of an analyte  is known to gradually change over time (Table 2). The DDT residues below the LLD were reported as not detected (nd).

Data analysis

The data obtained for the levels of all the six isomers of DDT, the mean total DDT (∑DDT) as well as the significant values between the DDT levels from the control and the exposed areas were calculated using Excel for mean and standard deviation or standard error values and for a one-tailed t-test. Standard deviation was used in the quality control data analysis while standard error was used for sample extracts data analysis. The level of significance was set at (5%). The DDT/DDE ratio was calculated to inspect the presence of DDT parent molecule as compared to its metabolites (Gonzalez et al., 2003).

Levels of DDT residues in indoor soil/dust samples

The levels of DDT and its metabolites detected in the indoor dust samples from the sprayed houses selected from villages of Apac and Oyam districts and the control Gwara village, Kiryandongo district are summarized in Table 3.

All the indoor dust samples collected from selected villages of Apac and Oyam districts contained detectable residues of DDT and its metabolites, while DDE and DDD residue isomers were above detection limits in 50% of the total indoor dust samples from the control village of Gwara. The levels of ∑DDT detected in indoor soil dust samples collected from Apac and Oyam district were 69.9 and 113.4 µg kg^-1 dry weight (dw), respectively, while 1.13 µg kg^-1 dw from Gwara village. The levels of ∑DDT detected in indoor dust samples from selected villages of Apac and Oyam districts were significantly higher (P < 0.05) than those from Gwara village. The significant difference in the ∑DDT levels confirms that high DDT residues levels such as p,p’-DDT are as a result of the recent application of DDT insecticide in the houses during IRS in the villages of Apac and Oyam districts in 2008. The ratios of the mean levels of p,p’-DDT to p,p’-DDE (4.0)  for villages of Apac, and that of p,p’-DDT to p,p’-DDD (1.8) for villages of Oyam districts  also  justifies  the recent spraying of DDT, since such ratios greater than one (1) suggest recent application of DDT in the environment (Gonzalez et al., 2003). The results therefore, raise concerns regarding human exposure to the contaminated indoor dust, particularly the exposure to children during play activities and other behavioral characteristics such as ingestion of non-food items and regular hand-to-mouth contact (Roberts and Ott, 2006; Hwang et al., 2008; Van Dyk et al., 2010). The results also raise concern regarding the possible presence of the insecticide residues in indoor air (Manaca et al., 2012) which could lead to DDT exposure through inhalation by the occupants.

Levels of DDT residues in outdoor soil samples

Levels of DDT residues and its metabolites detected in soil samples collected around the sprayed homesteads from selected villages of Apac, Oyam districts and from the control areas are summarized in Table 4.

The presence of the DDE and DDD residue isomers in the control sites could have been due to the continuous degradation of the parent DDT isomers to their metabolites following the past spraying of DDT in Uganda to control agricultural pests in the 1960s (Wasswa et al., 2011). The past spraying of DDT in the control sites is supported by the ratios of p,p’-DDT to p,p’-DDE and p,p’-DDD (Table 4) which are all less than one and such ratios suggest a previous application of DDT in the environment (Gonzalez et al., 2003). The presence of the residue isomers of DDE and DDD in soil samples collected from KWR may also be attributed to atmospheric transport of the residues from other places where DDT had been used (Ssebugere et al., 2010) since the wildlife reserve was gazetted in 1964 (Bahengana et al., 1997) and no practices such as agriculture, settlement, that could have been the source of DDT is allowed in the area. From other studies, Ssebugere et al. (2010), Hellar-Kihampa (2011) and Muhugija et al. (2017), reported total DDT up to 50 μg/kg dw, ranging from 22.6 ± 2.1 to 1452 ± 1.1 μg/kg dw and 7.5 to 564.2 µg/kg dw in soils from South-Western Uganda, sugarcane plantation in Tanzania, and Eastern Lake Tanganyika basin in Tanzania, respectively which supports the fact that DDT accumulate in soils.

Levels of DDT residues in leafy vegetables

The levels of DDT residues and its metabolites are summarized in Table 5. The results show that leafy vegetables are contaminated with DDT residues. In villages from Apac district, the ∑DDT detected in S. lycopersicum, S. melongena, H. sabdariffa, A. esculentus, and G. gynandra samples were 1.7, 2.9, 2.1, 8.3 and 4.4 µg kg^-1 fresh weight (fw), respectively. All residues of DDT isomers were detected in A. esculentus and G. gynandra, residues of DDE isomers were detected only in S. lycopersicum and H. sabdariffa while those of o,p’-DDT and p,p’-DDE isomers were below detection limits in S. melongena samples. In leafy vegetables from selected villages of Oyam district, the respective levels of ∑DDT detected in S. lycopersicum, S. melongena, H. sabdariffa, A. esculentus, and G. gynandra samples were 10.3, 2.3, 3.4, 4.1 and 11.4 µg kg^-1 fw. All DDT isomers were detected in G. gynandra while only residues of DDE and DDD isomers were detected in S. lycopersicum and H. sabdariffa vegetable samples. In Gwara village (control site), only DDE and DDD isomers were above detection limits with the highest mean level in µgkg^-1 fw of 12.2 for o,p’-DDD in S. lycopersicum samples and the lowest level being 0.4 for o,p’-DDD in H. sabdariffa samples. The highest and lowest  mean  ∑DDT  levels  in  vegetable  samples  from  Gwara village were 5.8 µg kg^-1 in S. lycopersicum and 0.5 µg kg^-1 in A. esculentus.

The vegetable samples from the villages in the study districts were mainly contaminated with the insecticide residues in the order DDE > DDT > DDD. G. gynandra, A. esculentus, and S. melongena were the most contaminated vegetables in both study areas and the control site, probably because the plants have abilities to absorb the residues into their tissues in a short time (Gitari et al., 2018). The contamination of these plants is likely to be as a result of root uptake of DDT residues and its metabolites present in the soil, as opposed to aerial precipitation of DDT during IRS (Van Dyk et al., 2010). The high mean levels of DDE and DDD detected in vegetable samples from the study areas were not significantly different (p > 0.05) from the mean levels of DDE and DDD in vegetable samples from the control site, probably because similar plant tissues are capable of absorbing insecticides (Gitari et al., 2018; Ngabirano and Birungi, 2021) at the same rate. ∑DDT levels from the current study were however lower than those reported by Van Dyk et al. (2010) of 43 μgkg^-1 in vegetables from Vembe district S. Africa, two months after IRS of DDT.

Ingestion is the most possible route for human exposure to DDT in non-malaria areas (Tao et al., 2009). Studies have shown that ingestion through diet is responsible for 94.9% of the total uptake of DDT while 5.1% is attributed to inhalation and dermal contact (Guo et al., 2005). This means that since the vegetables are contaminated with DDT residues and its metabolites probably due to the IRS process, the people of villages of Apac and Oyam districts may be exposed to DDT residues and its metabolites via ingestion, dermal and inhalation pathways. Vegetables such as S. lycopersicum, S. melongena, H. sabdariffa, A. esculentus, and G. gynandra are among the main diet of the people of villages in Apac and Oyam districts. Results of analysis of the vegetable samples from the selected villages of the two districts indicated that these plants contain levels of DDT and its metabolites especially p,p’-DDE and o,p’-DDE. High levels of p,p’-DDT as compared to p,p’-DDE and p,p’-DDD, was detected in S. melongena, A. esculentus, and G. gynandra samples and the ratios of p,p’-DDT to p,p’-DDE were up to 144.1 and 1.8 for villages from Apac and Oyam districts, respectively. On the other hand, the ratios of p,p’-DDT to p,p’-DDD for S. melongena, A. esculentus, and G. gynandra samples vegetable samples were recorded to be up to 38.4 from villages of Apac district and 1.8 in Oyam district. According to Gonzalez et al. (2003), such high ratios confirm accumulation of the pesticide residues in these vegetables as a result of the recent spraying of DDT in the two districts.

The baseline study conducted by RTI (2008), in which crops including vegetable samples from Apac and Oyam districts were analysed, revealed that DDT residues were below detection limits. The  current  study  has,  however,  revealed that twelve years after IRS, the mean ∑DDT levels in the selected food crops were high (ranged between 1.72 and 11.4 µg kg^-1). Nevertheless, the mean level of ∑DDT in the vegetables were significantly lower (P < 0.05) than the Maximum Residue Limit (MRL) set by European Union (EU) for DDT and its metabolites in vegetables like S. lycopersicum, S. melongena, H. sabdariffa, and A. esculentus (0.05 to 0.5 mg kg^-1) (EU Pesticide Database, 2020). The levels therefore do not pose a health threat to the consumers of the vegetables. However, since DDT and its metabolites are bio-accumulative (Agustina et al., 2008), long term health effects may arise due to DDT bio-accumulation, and over consumption or long-term exposure may lead to health effects especially in pregnant women and children (Nnamuyomba et al., 2014; Ssemugabo et al., 2022). Studies have reported that even in small doses, DDT has numerous impacts on the environment and human life (Birnbaum, 2002) since man is at high trophic level in the food chain. Therefore, the people in the study villages of Apac and Oyam districts may be at risk with time due to the presence of DDT and its metabolites in the vegetables that they feed on, though cooking foods could reduce the amount ingested (Van Dyk et al., 2010) which in turn reduces its risks to humans (Muhugija et al., 2017).

In this study to assess the levels of DDT and its metabolites in indoor and outdoor environmental samples after the 2008 use of DDT in Malaria Control Program in Northern Uganda, all DDT isomers and the metabolites were above detection limits in indoor dust and outdoor soil samples collected from villages of Apac and Oyam districts except o,p’-DDT and  o,p’-DDD which was below detection limits in some indoor dust and outdoor soils, respectively from villages of Oyam district. The ∑DDT residue levels detected in the indoor dust samples from Apac and Oyam districts were significantly higher than those from the unsprayed areas. Most of the leafy vegetables were also contaminated by DDT residues. The findings suggest that IRS of DDT conducted in 2008 led to an increase in the ∑DDT levels in the house indoor dust/wall, outdoor soils and leafy vegetables. The results however do not raise concern regarding the potential health effects of DDT to the consumers of the vegetables since the levels were significantly lower than the EU EMRL. However, the shift of the ∑DDT from below detection limits of the residues in the baseline study to detectable levels reported in this study raise health concerns due to potential bio-accumulation of the insecticide. The findings in this study therefore, calls for  the  re-evaluation  of the necessary safety regulations that forms part of the IRS protocol especially at national level, to prevent unnecessary contamination of the immediate domestic environment. It is also important to note that countries still facing the problem of malaria should seek for other measures such as Ecohealth to deal with malaria vectors rather than IRS using DDT because it does not render the environment free from being contaminated by the insecticide and its residues.

The authors have not declared any conflict of interests.