Human Health Risk Assessment Through the Detection of Organochlorine Pesticides in Vegetables and Fruits from Dhaka, Bangladesh by Gas Chromatography Coupled to Mass Spectrometry (GC-MS)

This study investigated the levels of Organochlorine pesticides (OCPs) in vegetables and fruits from local markets of Dhaka, Bangladesh using gas chromatography coupled to mass spectrometry (GC-MS/MS) and analyzed the possible health risks of peoples through consumption of those fruits and vegetables. A total of 100 vegetables and 100 fruits samples were analyzed and 45% of vegetable samples and 40% of fruit samples were found to be contaminated with OCPs residues. Aldrin, p,p-DDE, Cis-Chordane, p,p-DDT, Endrin, α-Endosulfan and Lindane-I were detected in vegetable samples while Aldrin, p,p-DDE, p,p-DDT, α-Endosulfan and Lindane-I were detected in fruit samples. The HI values of Aldrin, p,p-DDE, Cis-Chordane, p,p-DDT, Endrin, α-Endosulfan and Lindane-I in vegetables were 0.109, 0.006, 0.396, 0.008, 0.153, 0.882 and 0.019 respectively while HI values of Aldrin, p,p-DDE, p,p-DDT, α-Endosulfan and Lindane-I in fruit samples were 0.0155, 0.0005, 0.0002, 0.0339, 0.0019 respectively. The average concentration of detected OCPs residues were below the FAO’s recommended maximum residue limits (MRL). It is therefore concluded that, even though OCPs residues are within the maximum tolerance range but continuous accumulation in consumer body through consumption of such vegetables and fruits may lead to chronic effects that could be fatal. detected in grape fruit (0.93±0.47 µg/kg), maximum p,p-DDE content was detected again in water melon (2.25±1.07 µg/kg) with minimum detected in guava (0.38±0.21 µg/kg), maximum p,p-DDT was detected in water melon (0.61 ±0.29 µg/kg) with minimum detection in banana (0.328 ±0.18 µg/kg), maximum β-Endosulfan was detected in banana (0.89±0.39 µg/kg) with minimum detection in lemon and maximum Lindane-I was detected in banana (1.03±0.79 µg/kg) with minimum detection in papaya (0.53±0.32 µg/kg). Further, Aldrin content was not detected in mango, litchi, guava, orange and lemon samples, p,p-DDE was not detected in mango, litchi, guava, orange, papaya, pineapple and lemon samples, p,p-DDT content was not detected in mango, litchi, guava, orange, papaya, grape fruit, pineapple and lemon samples, β-Endosulfan was not detected in mango, litchi, guava, orange, papaya and grape fruit samples and Lindane-I was not detected in mango, litchi, guava, orange and lemon samples. 0.396, 0.008, 0.153278, 0.882 and 0.0198 respectively. The HI values of Aldrin, p,p-DDE, p,p-DDT, α-Endosulfan and Lindane-I in fruit samples were 0.0155, 0.0005, 0.0002, 0.0339, 0.0019 respectively. Risk analysis shows that, there have no potential health risks through consumption of vegetables and fruits from those areas of Bangladesh but long term exposure can cause serious health problems. The HI values of detected OCPs are within the range of safe (gure 2). can exposure OCPs from other sources like contaminated sh, milk, eggs, meat etc. In summary, our study indicated that, there have no potential health risks through consumption of vegetables and fruits from the studied area. In this study, HI values of OCPs were lower than the risk level. This calculation was based on the national per capita consumption. This HI values may vary if there have difference in the per capita consumption rate in any particular area.


Introduction
In agricultural productions, pesticides are using worldwide to resist pest attacks and prevent diseases. The excessive use of pesticides is the cause of spreading environmental contamination in developing countries (Kaushik et al. 2009;Pérez-Serradilla et al. 2010). Recently to detect the traces of pesticide residues in food becoming the primary concern in pesticide research, to obtain a comprehensive assessment of food quality and to avoid potential health problems (Araoud et al. 2007).
OCPs are a broad category of extremely long-lasting contaminants. After being applied for a long time, they remain in the environment. Many OCPs are endocrine disruptors, which means they have a mild harmful in uence on the body's hormonal systems, resulting in adverse health outcomes (Mansilha et al. 2010). OCPs are a signi cant source of worry worldwide due to their negative eco-toxicological consequences. OCPs are persistent organic The majority of OCPs have been outlawed due to their great persistence as insecticides, yet their residues continue to pollute food and the environment (Abou-Arab 1999). OCPs have been discovered in both raw and cooked foods. Permissible levels of persistent organic pollutants type OCPs in foods are getting stricter as people become more aware of the hazardous risks they represent to humans (Keikotlhaile et al. 2010). For acute and chronic exposure to OCPs, carcinogenic, neurological, endocrinal, natal and neonatal disorders are growing (Williams, 2013). OCPs, including hexachloro cyclohexanes (HCHs) and dichloro diphenyl trichloroethanes (DDTs), were widely utilized in agriculture, industry, and public health initiatives late 1940s (Fang et al. 2017). DDT breakdown products or metabolites include DDD and dichloro diphenyl dichloroethylene (DDE). Human faced preterm birth and spontaneous abortion due to DDT exposure. At high plasma DDE concentration, one nested case-control study showed a fourfold increased risk of breast cancer in the United States. In 1970, Sweden rstly banned OCP for agricultural use and then extended it to other countries (Turusov et al. 2002).
Pesticides are employed in agricultural areas to reduce crop loss by eliminating the physiological activities of target organisms. OCPs can contaminate water, soil, and air because of their lipophilicity, bioaccumulation, long half-life, and long-range transportability. Pesticides are harmful for living organisms in the aquatic ecosystem such as plants, shes, microorganisms and invertebrates (Brown and Heimbach, 2003;Grande et al., 1994;Castillo et al., 2006;DeLorenzo et al., 2001;Frankart et al., 2003). Because of their low cost and activity against a variety of organisms, pesticides including DDT, HCH, aldrin, and dieldrin are utilized in most Asian developing countries (Fry, 1995;FAO, 2005;Gupta, 2004). As a result, pesticide residues may damage nal consumers, particularly when these commodities are consumed fresh. Pesticide residues on agricultural commodities are known to be carcinogens and/or toxins, therefore reducing their total daily intake are preferred. Maximum residue limits (MRLs), which are de ned by each country and can lead to controversies since residual levels that are acceptable in one country may be unacceptable in another, are regulated by maximum residue limits (MRLs). The required rates of application may range from country to country and within regions within the same country, depending on agricultural and climatic factors (Torres et al. 1997;Lawal et al. 2019). Pesticide residues on treated crops are being closely monitored and controlled in many countries.
Analytical procedures are required to screen, quantify, and con rm pesticide residues in fruits and vegetables for both research and regulatory applications. Pesticide residues in fruit and vegetables have been con rmed by GC-MS (Fernandez-Alba et al. 1994; Chen and Wang 1996;Gelsomino et al. 1997;Columé et al. 2001a, b). Gas chromatography is used in conjunction with either electron capture detection (ECD) or mass spectrometry (MS) in the majority of OCP determination procedures.
Our work aimed to assess the amount of OCPs in ten types of vegetables and fruits from Dhaka's local market. In this determination, we used gas chromatography coupled to mass spectrometry (GC-MS/MS). This method is a sensitive and selective methodology for determining and identifying a wide range of pesticides at the trace level. Knowing the amount of OCPs in fruits and vegetables, we can quickly analyze the possible risks to human health after consumption of those vegetables and fruits.

Sampling
Various fruits and vegetables samples were purchased from largest local market for fruits and vegetables in Dhaka named Kawran Bazar. Sampling was done between February, 2021 to May, 2021. The samples were taken to the lab and stored at 4°C until further analysis. The vegetable and fruit samples were extracted and analyzed for the detection of OCPs within 24 hours after collection. Samples were collected on three replicates and a total of 10 types of fruits and 10 types of vegetables were collected randomly. The vegetables analyzed for the present study were cabbage (Brassica oleracea var capitate), cauli ower (Brassica oleracea var botrytis), Radish (Raphanus sativus), Hyacinth bean (Lablab niger), Sweet gourd (Cucurbita maxima), Cucumber (Cucumis sativus), Bitter gourd (Momordica charantia), Brinjal (Solanum melongena), Tomato (Lycopersicon esculentum) and Chilli (Capsicum species). The fruits analyzed for this study were Banana (Musa sapientum), Mango (Mangifera indica), Litchi (Litchi chinensis), Papaya (Carica papaya), Guava (Psidium guajava), Orange (Citrus chrysocarpa), Grape Fruit (Citrus grandis), Pineapple (Ananas comosus), Watermelon (Cucumis melo) and Lemon (Citrus limon). This research was conducted at the Institute of Food Science and Technology (IFST), Bangladesh council for Scienti c and Industrial Research (BCSIR), Dhaka, Bangladesh.

Extraction of sample and clean-up
Sample extraction was done according to the method (Takatori et al. 2011) with slight modi cation. In short, frozen samples were homogenized and thawed to room temperature. About 10 g of sample was transferred to 50 ml polypropylene tube and 10 ml of acetonitrile was added with it. The samples and the solution were homogenized with vortex mixer at high speed for 1 minute. Then, 4 g anhydrous magnesium sulphate and 1g of sodium chloride was added and shaken for 1 minute. Then, centrifuged for 5 minutes and 4 ml of organic layer supernatant was collected and eluted with two portions of 5 ml acetonitrile in a 50 ml ask (pear-shaped). The elute was then evaporated using a rotary evaporator. The concentrated extract was dissolved in 2 ml ethyl acetate and nally transferred to 2 ml standard vial for GC-MS/MS analysis.

Instrumentation and Gas Chromatography (GC)
The analysis was performed in GC-MS (Model: TRACE 1310, Thermo Fisher Scienti c, USA) equipped with Thermo Scienti c™ Trace GOLD™ TG-5MS GC Column (0.25 mm X 0.25 µm X 0.25 m) and 5% phenyl phase. Helium was used as the carrier gas in this experiment, with a constant ow rate of 1.2 ml/min. The injection port had a temperature of 230 degrees Celsius. The temperature pro le for GC was ranged from 80•C to 290•C. The inject volume was 2 µL. Spectral detection was performed by Mass Spectrometer (Model: TSQ DUO, Thermos Scienti c, USA).

Linearity of Calibration curve
Before starting the analysis, the column performance was checked by running only the blank samples. The stock solution of OCPs pesticides containing a cocktail of 19 pesticides was used as standard solution. For the standard calibration curve construction, ve different concentrations ranging from 5 ppb to 200 ppb were prepared and injected on the column and all the standards showed a linear range from 5 ppb to 200ppb. The coe cient values (R 2 ) obtained was ranged from 0.94 to 0.99 for all 5 standards.

Evaluation of Recovery performance
Recovery performance evaluation was used to con rm the method's precision results. For purpose of the recovery performance evaluation, each pesticides was spiked in distill water with two known concentrations (0.5µg/L and 1 µg/L). The spiked samples were extracted and analyzed exactly using the same procedure of sample analysis. The following equation was used to determine the mean percentage recoveries of the OCPs: Here, Pi represents the percent recovery, Si represents results from laboratory controls and Ti represents the percentage recovery from the spiked samples of known concentrations. Recovery tests were repeated at least two times for each OCPs and the mean percentage value, standard deviation (SD) was detected and presented in the table 2.

Determination of limit of detection (LOD) and Limit of quanti cation (LOQ)
Average blank value method was applied to determine the limit of detection (LOD) and limit of quanti cation (LOQ). LOD and LOD values were determined by running several blank samples. The LOD value of the OCPs was calculated based on the signal to noise ratio (3:1). LOD is de ned as the lowest concentration of OCPs that produced a chromatographic peak that was 3 times larger than the noise background (during the same retention time). The 10 times value of baseline noise in the chromatogram of the blank samples was used to compute the LOQ value.
2.8 Assessment of human health risks 2.8.1 Estimated daily intake (EDI) of OCPs According to the Codex Alimentarius Commission Procedural Manual (Codex Alimentarius Commission, 2006), the exposure assessment can be de ned as "the qualitative and/or quantitative evaluation of the likely intake of biological, chemical, and physical agents via food, as well as exposures from other sources if relevant." For each pesticide exposure the life time exposure dose (mg/kg/day) can be obtained through multiplication of residual pesticide concentration (mg/kg) in food items with the daily food consumption rate (per capita consumption) (kg/day) and dividing the result by the body weight (kg). The per capita consumption of vegetables and fruits in Bangladesh are 167.30 g/person/day and 35.78 g/person/day respectively (BBS 2017). Dietary exposure was determined by the following formula:

EDI = (FCC ×DFC) /BW
Here, EDI represents Estimated daily intake (EDI) (mg/(kg/day)); FCC indicated food chemical concentration (mg/kg); DFC is stand for daily food consumption (kg/day) and BW is used for Body weight (kg). The hypothetical assumption of U.S. Environmental Protection Agency -USEPA (1986,1989) about the body weight of adults is 70 kg considered for the dietary exposure calculation.

Hazard Risk Index (HI)
Consumers' health risk assessment from pesticide-contaminated vegetables and fruits was expressed as a Health Risk index (HI). The HI was calculated by multiplying the estimated daily intake (EDI) by the corresponding appropriate daily intake (ADI) values as given by WHO/FAO (Australian Government Department of Health, 2016) and indicated by the equation:

For vegetables,
HI v = Estimated daily intake of vegetables (EDI V )/ADI v Here, EDI v stands for EDI of vegetables and ADI v stands for acceptable daily intake of vegetables.
For fruits, HI f = Estimated daily intake of fruits (EDI F )/ADI f Here, EDI f stands for EDI of fruits and ADI f stands for acceptable daily intake of fruits.
If the value of HI is greater than 1, then it can be said that, the population consuming OCPs containing fruits and vegetables are safe from any health hazards.

Data analysis
The obtained results were summarized, organized, tabulated and analyzed through Microsoft O ce Excel 2007 and Statistics 10 (for statistical analysis). Duncan's multiple-range test was used to identify the differences between variables (DMRT).

Recovery evaluation
For the method evaluation, recovery performance was carried out. The recovery rate was determined by the concentrations of the standard samples.
Both fruit and vegetable samples were spiked with known two concentrations: 0.5 µg /Kg and 1 µg/Kg. The spiked samples were compared with that of blank samples because, the real blanks samples never could be acquired without target samples. Table 2 represents the obtained recovery (%) of those pesticide residues by using this method. The recovery of the OCPs was varied from 87.2% to 99.27% for vegetables and 89.54% to 100.3% for fruits samples respectively. The Chromatogram of the standard mixture containing 19 OCPs is presented in gure 1.

Determination of Limit of detection (LOD) and Limit of quanti cation (LOQ)
LOD was detected after injecting standard samples of three different concentrations at least six times of each sample. Then, mean standard deviation (SD) was determined for each residue. Figure 1 illustrates the chromatogram of standard samples. The LOD values were detected in the range of 0.019 to 0.033 µg/kg for vegetable samples and 0.017 to 0.038 µg /kg for fruit samples respectively. The LOQ values were detected in the range of 0.048 to 0.081 µg/kg for fruit samples and 0.049 to 0.088 µg /kg for vegetable samples. Table 1 shows the Retention time of each standard pesticides, LOD and LOQ values.

Detection of OCPs residues in vegetables and fruit samples
Total 100 samples from each category (vegetables and fruits) were analyzed. Mean residue levels of OCPs in vegetables and fruits are depicted in tables 3 and 4 respectively. Almost all of the samples were contaminated with OCPs residues. However, none of the sample contained residues of any OCPs above maximum residue limits FAO/WHO (1996) xed maximum residue limits (MRL).
Among the fruits samples, maximum Aldrin was detected in water melon (3.67±1.98 µg/kg) with minimum detected in grape fruit (0.93±0.47 µg/kg), maximum p,p-DDE content was detected again in water melon (2.25±1.07 µg/kg) with minimum detected in guava (0.38±0.21 µg/kg), maximum p,p-DDT was detected in water melon (0.61 ±0.29 µg/kg) with minimum detection in banana (0.328 ±0.18 µg/kg), maximum β-Endosulfan was detected in banana (0.89±0.39 µg/kg) with minimum detection in lemon and maximum Lindane-I was detected in banana (1.03±0.79 µg/kg) with minimum detection in papaya (0.53±0.32 µg/kg). Further, Aldrin content was not detected in mango, litchi, guava, orange and lemon samples, p,p-DDE was not detected in mango, litchi, guava, orange, papaya, pineapple and lemon samples, p,p-DDT content was not detected in mango, litchi, guava, orange, papaya, grape fruit, pineapple and lemon samples, β-Endosulfan was not detected in mango, litchi, guava, orange, papaya and grape fruit samples and Lindane-I was not detected in mango, litchi, guava, orange and lemon samples. As we know, several OCPs are banned in some developed countries. The OCPs residues detected from the vegetable and fruit samples may be due to the sporadic use of these pesticides in agricultural elds or may be past extensive use of these pesticides for agricultural production. From this analysis it is also found that some of the vegetables and fruits were contaminated with more than one OCP. There may have several reasons for this such as farmers may mix more than one pesticides at the same time of application. Other possible reasons may be contamination by water, soil or air. As the vegetables and fruits have similar contamination ranges therefore it suggests that those vegetables or fruits were contaminated from almost similar sources.
Although the concentration of pesticides residues are below the FAO/WHO recommended MRL but continuous exposure of OCPs residues through consumption of those vegetables and fruits can accumulate in the consumers body and later may develop chronic effects in human population for long term. The study also supports that the farmers specially the vegetables growers are using those pesticides which is responsible for the contamination of those vegetables. Fruit cultivars might also use pesticides for better production or those fruits may be contaminated from the other vegetables or fruits.
This study also suggests that farmers from those regions are not properly following good Agricultural practices (GAP). Farmers should be trained properly to enhance their knowledge on pests, their ecological casualties, non-chemical alternatives and proper management of pesticides.

Health risk from consuming these fruit and vegetable samples
The EDI (Estimated daily intake) and HI (Health risk index) values analyzed from this study are presented in Table 5 and 6 for vegetables and fruits respectively. Theoretically, if Health risk index is greater than 1, then the responsible pesticide may cause potential health risks for systemic effects (Thornburg 1973;Tchounwou et al. 2002;Reffstrup et al. 2010). The HI values of Aldrin, p,p-DDE, Cis-Chordane, p,p-DDT, Endrin, α-Endosulfan and Lindane-I in vegetables were 0.109, 0.006, 0.396, 0.008, 0.153278, 0.882 and 0.0198 respectively. The HI values of Aldrin, p,p-DDE, p,p-DDT, α-Endosulfan and Lindane-I in fruit samples were 0.0155, 0.0005, 0.0002, 0.0339, 0.0019 respectively. Risk analysis shows that, there have no potential health risks through consumption of vegetables and fruits from those areas of Bangladesh but long term exposure can cause serious health problems.
The HI values of detected OCPs are within the range of safe levels ( gure 2). Moreover, peoples can exposure OCPs from other sources like contaminated water, sh, milk, eggs, meat etc. In summary, our study indicated that, there have no potential health risks through consumption of vegetables and fruits from the studied area. In this study, HI values of OCPs were lower than the risk level. This calculation was based on the national per capita consumption. This HI values may vary if there have difference in the per capita consumption rate in any particular area.

Conclusions
In this study, OCPs are quanti ed in vegetables and fruits samples and assessed the potential human health hazards. Several OCPs such as Aldrin, p,p-DDE, Cis-Chordane, p,p-DDT, Endrin, α-Endosulfan and Lindane-I were detected in vegetable samples while Aldrin, p,p-DDE, p,p-DDT, α-Endosulfan and Lindane-I were detected in fruit samples. The detection range was within the safety limits. As we know, OCPs are non-biodegradable and thus remain in the environment as pollutants. Risk assessment is an important procedure to quantify the potential health risks and provides information the risk managers to control the overuse of OCPs. Hence, a safe environment ensures the food safety to the consumers. The current study indicates that, there may have several health hazards due to OCPs exposure through fruits and vegetables. In conclusion, the OCPs exposure must be controlled to reduce the human health risks. Although, our study provided estimation about the presence of OCPs in fruits and vegetables but, further studies are suggested to ensure the safety of consumers.

Declaration of Competing Interest
The authors declare that they have no known competing interests.