Estimation and human health risk assessment of organochlorine and organophosphate pesticide residues in raw milk collected in Kenya

Study area

Kiambu, Isiolo, and Laikipia County were selected as study areas. Kiambu County (-1⁰10’0.01’’ S-36⁰49’59.99’’E) has an area of 1679 m above sea level and is generally a dairy and tea zone area¹⁵. Sheep farming, horticulture and maize farming is also practiced. There are about 2.4 million inhabitants in Kiambu scattered over an area of 2450 km²¹⁶. The county currently leads in cow milk production in Kenya¹⁷.

Isiolo County (0⁰21’16.63’’N-37⁰34’55.85’’E) has a population of about 200,000 inhabitants and an area of 25,700 Km^216,18. The main economic activity is livestock production and nomadic pastoralism is the major lifestyle of the inhabitants of this county¹⁸. Laikipia County (0⁰01’0.01’’N-37⁰04’22.19’’E) has a population of about 520,000 people and an area of about 8700 Km². The map of the study area is shown in Figure 1.

Figure 1. Map of the study area.

Sample collection and preservation

The study was conducted between 2017 and 2018. Milk samples from 8 different wards of Kiambu, Isiolo and Laikipia were collected in 50 mL falcon tubes. A total of 90 cow milk and 82 camel milk samples were provided by farmers from collection centers for analysis. The samples were labelled (date, location, time), packed into ice-bags and transported to the Analytical Chemistry and Food Safety Laboratory-Directorate of Veterinary Services, Kabete for storage at -20 °C in the deep freezer before analysis.

Laboratory analysis of pesticide residues in milk

Chemicals and reagents. Acetonitrile and Isooctane of high purity (99.9%) were purchased from Merck. Organophosphate standards (methacrifos, disulfoton, terbufos, chloropyrifosmethyl, tolclofosmethyl, fenchlorphos, malathion, tetrachlorvinfos, fenamiphos, chlorthiophos, profenofos, leptophos, and coumaphos) and Organochlorine standards (Alpha- Benzene Hexachloride [BHC], beta-BHC, Hexachlorobenzene, gamma-BHC (lindane), Delta-BHC, heptachlor, cis-chlordane, trans-chlordane, p’p-DDE, Alpha-endosulfan, dieldrin, Op’-DDD, Trans-nonachlor, chlorfenson, p,p’-DDD, Endrin, Beta-Endosulfan, p’p-DDT, Op’-DDT, Cis-Nonachlor, and mirex) were purchased as a gas chromatography multiresidue pesticide kit (100 μg/mL each in toluene, 1mL/ampoule) from Restec company. Quick, easy, cheap, effective, rugged, and safe (QuEChERS) extraction tubes, preweighed salts (6 g magnesium Sulphate, 1.5g sodium acetate) and 15 mL fatty samples AOAC tubes with ready to use salts (400 mg PSA, 400 mg C18EC, 1200mg magnesium sulphate) were all purchased from Agilent Technologies through Nesvax Innovations Ltd Nairobi, Kenya.

Preparation of standard solution. A stock solution (1μg/mL) was prepared by weighing 200 μL of the multi residue pesticide standard into a 20 mL volumetric flask and toped up with isooctane. The standard stock solution was used to prepare various working concentrations through serial dilutions. The standard stock solution and other working solutions were stored in a refrigerator at 4°C. The standard for spiking experiment was obtained by transferring predefined volumes of family mixes from 150 μL–1500μL.

Extraction of pesticide residues in milk. Milk samples were extracted using the QuEChERS technique. Briefly, 15 mL of whole milk was weighed into a 50 mL polypropylene centrifuge^19,20. An appropriate amount of the pesticide spiking standard was added to the fortified blank sample and vortexed for 60 seconds, allowing 1 minute for the interaction of the matrix and the standard before extraction^19,20. Acetonitrile (15mL) was added to each milk sample, capped, vortexed for 1 minute and pre-weighed salts (6g magnesium sulfate, 1.5g sodium acetate) were added and immediately shaken vigorously by hand for 1 minute to prevent the agglomeration of salts. The resultant mixture was centrifuged at 5000 rpm for 10 min at 4°C^19,20, 6 mL of the supernatant layer were then transferred into a 15 mL tube containing 400 mg PSA, 400 mg C18EC, and 1200 mg magnesium sulfate. The resultant mixture was capped and vortexed for 1 minute then centrifuged at 5000 rpm for 10 minutes^19,20. Following this, 1 mL of the extract was transferred into another tube, dried by nitrogen flow at 40°C, reconstituted with 1 mL of isooctane and filtered through disposable 0.45μm membrane filters into gas chromatography tandem mass spectroscopy (GC-MS/MS) auto sampler vials for GC-MS/MS analysis.

GC-MS/MS. The pesticides residues were detected using a GC-MS-TQ8040 (Shimadzu, Japan) machine equipped with Shimadzu manufactured AOC-20s auto sampler and AOC-20i auto injector. The analytical capillary column was a ZB-5ms with a thickness of 0.25 μm, a length of 30.0 m, and a diameter of 0.25 mm. The column temperature was maintained at 50 °C for 1 minute then programmed at 25 °C/minute up to 125 °C, then finally 10 °C up to 300 °C for 3.50 minutes. This was held for 20 minutes. Helium (99.999% purity) was used as the carrier gas at a flow rate of 1.69 mL/min. The injection port temperature was 250°C and 1 μL was used as the injection volume with the spitless mode/purge flow was 5 mL/min. The ion source temperature was 200 °C, the interface temperature was 250 °C and the solvent cut was 1.5 minutes with a detector voltage 0.5 kV. Multiple reaction monitoring (MRM) was used to convert MS1 (parent ion) to MS2 (daughter ion). The retention time windows and base peak ions were designed for OCPs and OPPs. All pesticides were identified with specific ions and retention time and were quantified using the external standard method.

Quality assurance

Recovery studies were carried out using blank milk samples spiked at 10 ppb, 50 ppb and 100 ppb to determine the accuracy of the method. The amount of analyte which was recovered after complete sample extraction and processing was determined. The average recoveries of spiked OCP samples ranged from 73% to 97% and 78% to 100 % for OPPs. To determine the linearity of the method, a standard curve was constructed. Calibration curves of organophosphates and organochlorines were prepared using concentrations of 20 ppb, 30 ppb, 50 ppb, 100 ppb and 200 ppb. The calibration standards were prepared from a 1 ppm stock solution by using serial dilutions of the mixed OCPs and OPPs. Linearity was determined by regression analysis of the peak area against the concentration of the analyte and was expressed using the linear regression coefficient (R²). Generally, an R² value of >0.998 is considered as evidence of an acceptable fit of data to the regression line. In this study the linear regression range was from 0.991–0.999.

The ability of the method to detect and separate OCPs and OPPs in the mixture without interference from other components in the mixture was determined. The selectivity of the method was determined when chromatographic peaks obtained from the mixture of the pesticide standard analysis showed the absence of interfering peaks. The lowest concentration at which OCPs and OPPs could be detected and also quantified was determined using the formula S_LOD=S_RB+3σ_RB, SLOQ=S_RB+10σ_RB, where S_LOD is the signal at the limit of detection (LOD), S_LOQ signal at the limit of quantification (LOQ), S_RB signal of the reagent blank, and σ_RB standard deviation of the reagent blank. According to the directives of the International Union of Pure and Applied Chemistry (IUPAC) and the American Chemical Society’s Committee on Environmental Analytical Chemistry the LOD for OCPs range from 0.1–0.3 ng/mL, the LOQ for OCPs range from 0.1–0.4 ng/mL, the LOD for OPPs range from 0.02–0.3 ng/mL, and the LOQ for OPPs range from 0.03–0.4 ng/mL. The relative standard deviation (RSD) was less than 3%.

Human health risk assessment. Dietary exposure, in the form of estimated daily intake (EDI), and the chronic/long-term hazard quotient (cHQ) are a measure of the health risk due to ingestion of milk contaminated with OCPs and OPPs. These were calculated for adults (aged above 18 years) and children (3–10 years) using the formula described by John et al.²¹.

i.e. $\text{EDI}=\frac{\text{C}\ast \text{Fi}}{\text{Bwt}}$

A body weight of 70 kg was used for adults (aged above 18 years) and 23 kg was used for children (aged 3–10 years) as recommended by the European Food Safety Authority²². C represented the average concentration of the chemical contaminants (ng g/) in the milk, Fi was the average daily intake and was based on the information retrieved from United States Agency for International Development-Kenya Agricultural Value Chain Enterprises (USAID-KAVES) dairy value chain analysis for adults (380 g) and Ogenche et al. for children (250 g)^23,24. EDIs were compared with the acceptable daily intake (ADI) values^25,26. The cHQ was calculated by the formula cHQ=EDI/ADI²⁷.

Levels of OCPs

The recorded results revealed that the percentage of cow milk samples contaminated with OCPs ranged from 0.00 to 93.85% while the percentage of camel milk samples contaminated with OCPs ranged from 0.00-96.15%. These ranges were higher than those detected in goat (75%), buffalo (75%), and cow milk (50%) samples by Raslan and colleagues in Zagazig city in Egypt²⁹. However, the percentage of positive samples in cow milk reported in the current study were lower than the percentage of positive samples in buffalo and sheep milk marketed in Brazil (100%)³⁰. In general, the values of OCPs ranged from below the detection limit to 22.38 ppb and the mean values ranged from 0.00 to 12.38 ppb. These values were lower than what was reported in human milk from Tanzania, cow, buffalo, and goat milk from Egypt, and goat milk and cheese marketed in Ethiopia and Ghana^29,31–33. The results also showed that the detected concentrations of the OCPs were low when compared with the established maximum residue levels (MRLs) established by the Codex Alimentarius, the EU Pesticide Database and the USDA^34,35. However, the mean levels of heptachlor evaluated in cow and camel milk collected in the study area were above the maximum residue limits set by the Codex Alimentarius and the EU Pesticides Database. These findings are in agreement with those of Omwenga and colleagues who reported high levels of heptachlor in farmed fish collected in the Kiambu and Machakos counties¹⁵. Heptachlor is a persistent organic pollutant (POP) which has been reported to be present in soil for up to 14 years after its initial application³⁶. Like other POPs, it is lipophilic and poorly soluble in water and tends to accumulate in the body fat of humans and animals³⁶. This organochlorine was banned in the US in the 1980s and is considered a potential carcinogen by the International Agency for Research on Cancer and the Environmental Protection Agency (EPA)³⁶. Animals exposed to heptachlor during gestation and infancy were reported to have altered immune function, decreased body weight and some died³⁷.

BHC isomers and DDT were detected in some of the milk samples in the study area. The use of these pesticides was banned in Kenya on account of their persistence in the environment and toxicity to untargeted organisms¹⁵. Some of the isomers of BHC were initially used for seed dressing to protect crops against termites and for use in cattle dips. pp’-DDT was detected in 47.35% (Kiambu) of cow milk samples, and 30% (Isiolo) or 43.75% (Laikipia) of camel milk samples. These values were lower than the 50%, 75%, and 75% in cow, buffalo and goat milk samples in Egypt²⁹. The presence of these OCPs in milk samples suggest previous use of these pesticides in agricultural activities in Kenya similar to what was observed in previous studies in Egypt (milk), Ghana (milk, yoghurt, and cheese), and Ethiopia (milk)^29,31,32. Could farmers in the study area be using these pesticides illegally? Could their detection be related to previous application? These are pertinent queries that require further inquiry.

Human dietary intake of OCPs and health risk assessment

Humans may be exposed to OCPs in a number of ways including polluted air, dermal ingestion, or via contaminated food and/or water²⁹. The main source of human exposure to pesticides include foods such as milk and dairy products^29,38. In the current study, the human dietary intake, EDI and risk assessment (cHQ) of pesticides were calculated using ADI values previously summarized by Dourson and Lu^25,26. On the basis of this summary, the estimated daily intake and chronic/long-term hazard quotient values of only 11 out of 18 OCPs could be estimated. When cHQ values exceed 1, a health risk cannot be excluded²⁷. The results of cHQ suggest that the consumption of cow’s milk by adults and children is alarming for endosulphans, dieldrin, heptachlor and DDT and the consumption of camel’s milk by adults and children is alarming for endosulphans, dieldrin, and heptachlor. A similar trend was observed in the case of heptachlor and dieldrin in Egypt and DDT in Arusha, Tanzania^29,33.

Levels of OPPs

It was established that the mean levels of all the 14 organophosphates tested were below the maximum residue limits set by the Codex Alimentarius, EU pesticides database, and the USDA. It may therefore be plausible that i) farmers in the study area may be adequately educated on the use of these pesticides, ii) the use of these pesticides may be generally low in the study area, or iii) proper handling of milk is practiced in the study area which minimizes contamination of the milk with the pesticides. However, the observation that cow milk registered higher mean levels of some organophosphate pesticides, including Chlorthiophos-2, Coumaphos, Fenchlorphous, Leptophos, Malathion, Profenefos, and Tetrachlorvinphos, relative to camel milk may have something to do with the fact that Kiambu county (where the cow milk was collected) is an agriculture intensive area and there may be a possibility of prior use of these pesticides relative to Isiolo and Laikipia counties (where camel milk was collected) which are mainly inhabited by pastoralists who may not be practicing large scale agriculture. Leaching and surface run-off could be another potential reason for the high levels of these organophosphates³⁹.

In total, 15.34% of cow milk samples from Kiambu, 70.31% of camel milk samples from Isiolo and 43.75% of camel milk samples from Laikipia were contaminated with Malathion. This is in contrast to the 44% reported in raw buffalo milk collected in Assuit Egypt⁴⁰. In general, the values of OPPs ranged from below the detection limit to 387.47 ppb. The mean values ranged from 0.00 to 46.07 ppb, and the percentage of contaminated samples ranged from 0.00 to 93.18%. A previous study on the levels of selected organophosphates in human colostrum and mature milk samples in Uttar Pradesh reported high ethion contamination in colostrum (23.1% or 6/26) and high levels of chlorpyrifos contamination in mature milk samples (50% or 4/8)⁴¹. The study further reported that the percentage chlorpyrifos contamination was 19.2% (5/26), dimethoate contamination was 3.8% (1/26) while 25.0% (2/8) of mature samples were contaminated with dimethoate, and 12.5% (1/8) of mature milk samples was contaminated with ethion⁴¹. Another study in Pakistan evaluated the dietary transfer of pesticides to dairy milk and its effect on human health. The study reported that about 40% and 20% of cow milk samples recorded residue levels of cypermethrin+chlorpyrifos and profenofos that was higher than their maximum residue limit as suggested by the World Health Organization⁴².

The primary objective of organizations such as the World Health Organization, the United States EPA, the Codex Alimentarius, and the EU Pesticides Database is to develop MRLs that are aimed at protecting the health of consumers while facilitating trade⁴³. The MRLs are set at values which are not higher than those resulting from pesticide use in line with good agricultural practices (GAP)⁴³. Most developing countries have not developed their own internal regulatory mechanisms for pesticides choosing instead to rely on the aforementioned organizations for guidance. In this context, it was worrying to observe that 10/14 of the studied OPPs did not have MRLs as per the Codex Alimentarius, 6/14 of the studied OPPs did not have MRLs as per the European Pesticides Database, and 10/14 of the studied OPPs did not have MRLs as per the USDA. This lack of regulation may be a huge pitfall as far as monitoring the levels of these chemicals is concerned.

Human dietary intake of OPPs and health risk assessment

Organophosphates have a high acute toxicity to target organisms⁴. However, they are not as persistent in the environment as organochlorines because they decompose into non-toxic products⁴. Notwithstanding, on the strength of the analyzed data, it may be inferenced that the health of children in the study area may be at a higher risk from organophosphates than adults. Data from Africa on this phenomenon is scant but a similar finding was observed in a study in India on the pesticide residues in peri-urban milk from Bhubaneswar, Guwahati, Ludhiana, and Udaipur in India⁴⁴.

Underlying data

Figshare: Data for estimation and human health risk assessment. https://doi.org/10.6084/m9.figshare.16823788.v1²⁸

This project contains the following underlying data:

Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).