Estimating combined health risks of nanomaterials and antibiotics from natural water: a proposed framework

Nanoparticles (NPs) are one of the major class of emerging contaminants identified in aquatic environment. There is a probability that they can co-exist with other chemical pollutants like antibiotics (ABs) as ABs-NPs complexes in natural water systems. If these complexes are taken up via inadvertent ingestion of contaminated water, it might show detrimental effects on human health. To address this challenging issue, this study developed a risk framework to assess the combined exposure of ABs and NPs in natural waters for the first time. The six-step framework was applied to a hypothetical exposure of NPs (copper oxide, CuO; zinc oxide, ZnO; iron oxide, Fe3O4; and titanium oxide, TiO2) and ABs (ciprofloxacin, CIP; ofloxacin, OFX; norfloxacin, NOR; and levofloxacin, LEVO) to estimate human health risks for two different exposure scenarios. Risk estimation was also conducted for the released fragments of ABs, NPs and metal ions in the human digestive system. Mixture toxicity risk assessment was conducted for three different combinations: (i) ABs and metal ions, (ii) ABs and NPs, and (iii) NPs and metals ions. Although the expected risk values were observed to be less than 1 (both hazard quotients and hazard interactions less than 1) for all the conditions and assumptions made, still a thorough monitoring and analysis of the studied contaminants in water is required to protect humans from their adverse effects, if any. Maximum allowable concentrations (Cmax) at which no risk can occur to humans was found to be (maximum values): ABs (233.8 µg/L, NOR); metal ions (1.02 × 109 mg/L, Ti2+ ions), and NPs (6.68 × 105 mg/L, TiO2), respectively.


Introduction
In the past few years, increased concerns have been raised due to occurrence of a wide variety of emerging contaminants including nanomaterials and pharmaceutical drugs in Communicated by Lotfi Aleya. natural water systems. Studies suggest that the concentration of these contaminants in water ranged from µg/L to ng/L (Chen et al., 2016;Ebele et al., 2017) and even high concentration values have also been reported. Nanoparticles (NPs) usually display unique physical and chemical properties, and because of their inherent reactivity with other pollutants, NPs may act as a carrier and co-exist with other pollutants producing long-term environmental and health risks (Azizi et al., 2016;Wang et al., 2016).
Upon release and emission, NPs may interact with other water contaminants like antibiotics and ABs in the environment and may exist as ABs-NPs complexes leading to possible co-exposure (Naasz et al., 2018). Both these contaminants are likely to be present and co-exist in the aquatic environment (Lammel et al., 2019). Although studies have been conducted to identify the individual occurrence of these contaminants Kumar, 2020a, 2020b), scarce or restricted information is available on their co-existence in natural waters and subsequent risk exposure to human health if taken inadvertently (Coll et al., 2016;Holden et al., 2014). In real-life scenarios, they might differ in their toxicities or even might undergo transformations to produce transformation products. These transformation products under certain circumstances may show detrimental harmful effects, and thus this aspect needs to be addressed.
There is a probability that human beings might get exposed to mixtures of contaminants (Uwizeyimana et al., 2017). During the process of passage, it is possible that NPs or ABs (pharmaceutical drugs) can form NP-toxin complexes or ABs complexes or even produce inter-category complexes of two contaminants because of NPs large surface area (Zhu et al., 2011). Thus, it is important to know the method for evaluating environmental risk due to mixtures containing ABs-NPs. In recent years, the scientific community has put tremendous efforts to assess the eco-toxic potential of nanomaterials (Menard et al., 2011), but comparatively few studies have investigated the possible interaction with co-existing "traditional" environmental pollutants (Canesi et al., 2015;Hartmann and Baun, 2010;Naasz et al., 2018).
Looking into the potential adverse effects of these contaminants on human health, risk assessment studies have been conducted for some classes of contaminants, for example, NPs (Parsai and Kumar, 2020) and pharmaceutical drugs (ABs) Kumar, 2020a, 2020b) (Supplementary Table S1), but none of the reported studies (as per authors best knowledge) has tried to capture the interaction aspect linked with NPs and ABs in natural water systems. A lack of available guidelines and regulations adds to the existing problem and makes it even more difficult, if not properly taken care of. Therefore, it becomes imperative to study and identify the interaction of these contaminants in water so that guideline values can be formulated for exercising appropriate control measures.
This study aimed at proposing a framework to determine the risk exposure effects of NPs and ABs to the human digestive media or GI-tract during the inadvertent ingestion of contaminated water. Four widely detected NPs (ZnO, CuO, Fe 3 O 4 , and TiO 2 ) and pharmaceutical drugs, ABs (ciprofloxacin, CIP; ofloxacin, OFX; norfloxacin, NOR; and levofloxacin, LEVO) identified in the natural water were selected for this study. Figure 1 shows the flow diagram of the proposed framework for determining human health risk estimates due to the interaction of ABs and NPs in natural water due to the inadvertent ingestion of contaminated water. The six-step risk assessment approach have been used for single contaminants also (Kumari and Kumar, 2020a;Kumar et al., 2014). Briefly, this framework assumes that when these contaminants enters into the human body via oral route they might get disintegrate into their respective ABs and NPs within due course of time in the human digestive system. The released NP might further get dissolved into their respective ions, releasing free NP. Although, zinc, copper, and iron are essential elements for all living organisms, these metals can cause harmful effects if present in excessive quantities (Evangelou et al., 2007;Twining et al., 2005). Therefore, this approach also assumes the probability of risk exposure effects due to released ABs and released NPs in the human digestive system. Risk was also estimated due to free NP and free metal ions in human digestive liquid. The size, shape, charge, and surface area of NPs were not considered as the primary objectives which were to develop the framework and show its application. The effects of these factors on risk estimate can be studied in future studies.

Hazard identification
Hazard identification involves identifying the material of interest and collecting information for which risk evaluation needs to be done. To determine the suitability of the suggested framework, this study considered the hypothetical exposures of NPs and ABs for illustrative purpose. NPs (CuO, ZnO, Fe 3 O 4 , and TiO 2 ) were considered as they are used as antimicrobial agents and additives in consumer and health-care products. The risk assessment of selected NPs is essential as several research investigations have shown their adverse effects to human health (Croteau et al., 2014;Ye et al., 2018). Amongst the NPs selected, iron oxide NPs (Fe 3 O 4 , γ-Fe 2 O 3 , and superparamagnetic IONPs) have been extensively used for pharmaceutical applications (Ding & Guo, 2013;Namvar et al., 2014). CuO NPs might induce oxidative stress resulting in destruction of human liver cell (Shukla et al., 2013). ZnO is one of the most frequently detected NPs in surface water (Kurlanda-Witek et al., 2014) and has been reported be cytotoxic and harmful compared to other metallic NPs (Li et al., 2020).
To study the interaction of NPs with ABs, the most widely detected fluoroquinolones (FQs) ABs (CIP, OFX, NOR, and LEVO) in the environmental media were selected owing to their use in the treating pulmonary, urinary, and digestive infections. The selected ABs are also effective in treating wide range of pathogenic bacteria (gram-negative and gram-positive) and mycoplasmas (Peterson and Kaur, 2018;Sharma et al., 2017). CIP, NOR, and OFX are common ABs administered to humans (Hamed et al., 2018;Sharma et al., 2010;Thai et al., 2021). Amongst these, CIP is one of the most extensively used drug in the world (Sukul and Spiteller, 2007), and LEVO has been listed as one of the most essential medicines of human use by the World Health Organization (Hooper and Rubinstein, 2003). The environmentally occurring concentration (EC) of ABs and NPs is taken from published scientific literature. The information is provided as supplementary text Table S2 and S3.

Exposure assessment
This research study evaluated health risks due to the inadvertent ingestion exposure of ABs and NPs to children as they have been recognised to be the most sensitive subpopulation compared to adults (Preston, 2004). Health risk effects were estimated for two different scenarios (Fig. 2). Detailed information is provided below.
Under scenario 1, the study considered NPs for which the adsorption isotherm data was available in literature. Scenario 2 indicates the condition where the adsorption isotherm data for the selected NPs is not available in published literature. For scenario 1, risk was estimated for two different exposure routes: (1) risks due to absorption of ABs in the digestive media only and (2) risk due to absorption of NPs in digestive media. Under exposure route 2, two sub-routes were considered: (a) route 2a: risk due to exposure to metal ions after the dissolution of NPs in digestive liquid and (b) route 2b: risk due to the revised concentration of NPs in digestive media. For scenario 2, two cases, (i) no absorption and (ii) 100% absorption, were considered. The detailed information is provided in the sections to follow.

Risk estimation of NPs in the digestive media during GI-tract absorption when the isotherm data is available
It is believed that oral ingestion serves as the main and primary route for the probable uptake of NPs to target sites or structures followed by their dissolution in the human digestive system (Fig. 1). In risk assessment, dissolution of NPs in the human digestive system gives a realistic and accurate quantity of  (Zhong et al., 2017). It is observed that the dissolution of NPs led to the release of metal ions in the digestive system. Therefore, it is also important to estimate risks due to the residual concentrations of released metal ions and NPs in the human digestive system. Risk estimation of metal ions released after bio-assimilation of NPs in the digestive media was carried out in terms of the average daily dose of metal ions (ADD Mi, DM ) to the human digestive system (Eq. (1)). Similarly, risk due to revised concentration of NPs in the digestive media during GI-tract absorption was calculated as ADD R_NP, DM (Eq. (2)).
In these equations, C Mi is the concentration of released metal ions in the digestive media; BAF NP is the bio-assimilation potential of NPs; C NP, DM is the revised concentration of NPs in the human digestive system.

Risk estimation of ABs in the digestive media during oral administration when the isotherm data is available
In this case, concentration of NPs was assumed to be 1 mg/L. Concentration of ABs after adsorption by NPs (mg/g) is taken from published literature, (supplementary information, Table S2). To determine the concentration of ABs after absorption on NPs in mg/L, the concentration of NPs (mg/L) was multiplied by concentration of ABs after absorption on NPs (mg/g) and divided by 1000, which is the conversion factor. To calculate the concentration of ABs in the human digestive media during GI-tract absorption (Conc AB_DM ) (Eq. 3) after oral ingestion, dissolution rate of ABs in the digestive media and the rate at which ABs is getting absorbed by the GI-tract are taken into account. Dissolution rate and absorption rate of data of ABs are taken from those reported in literature and are provided as supplementary Table S4.
To determine the risk exposure effects, chronic daily intake (CDI) values (mg/kg-bw/day) were estimated using Eq. (4).
where, C AB_DM is the concentration of ABs in the digestive media during GI-tract absorption in GI-tract in mg/L, BW is the body weight (kg); IR is the intake rate of water, EF is exposure frequency (365 days/year), and ED is the exposure duration (70 years).

Risks estimation when the adsorption isotherm data were not available
This study also tried to determine risk exposure effects for a hypothetical situation where the adsorption isotherm data of ABs on NPs is not available. Under this scenario, two different cases, (i) no absorption and (ii) 100% absorption, were considered to determine risk estimates.
(1) No absorption: It is assumed that no absorption of NPs takes place in the GI-tract, and the ABs ingested remain as free ABs in the digestive media. Risk exposure of ABs was estimated using the surface water concentration of ABs as mentioned in the hazard identification section. Acceptable daily intake (ADI) values of individual ABs were used to calculate the predicted noeffect concentration (PNEC) values. ADI values specify the level of daily intake that should not result in any damaging effects to human health from direct exposure (Cunningham et al., 2009) whereas PNECs represent the lowest concentration values at which no harmful effects are anticipated. It provides more accurate estimates and can be refined using detailed information considering different assessment factors (Bopp et al., 2019). Input parameters used to estimate PNEC values and ADI values of the ABs are given in Table 1. where, ADI is acceptable daily intake (µg kg-day −1 ); BW is body weight of children in kg; IR w is the intake rate of water in L day −1 ; and GI AF is the gastrointestinal absorption factor of ABs, assumed to be 1 in this study.
(2) 100% absorption: Under this case, it is assumed that the amount of NPs and ABs is getting fully absorbed in the GI-tract. Risk was estimated similar to the approach mentioned in "Risk estimation of NPs in the digestive media during GI-tract absorption when the isotherm data is available"section.  (Parsai and Kumar, 2020). ADI values of individual ABs were taken from published literature. Table 1 lists information about the parameters used for risk estimation.

Risk estimation and characterisation
Hazard quotient (HQ) values were calculated to determine risk to children. HQ is the ratio of the possible exposure to a contaminant and the extent to which no adverse effects are expected to occur. If the HQ is observed to be less than 1, then no adverse health consequence is anticipated as a result of exposure (Kumari et al., 2015). HQ values were calculated using the (i) CDI and ADD values estimated in the exposure assessment section and (ii) RfD values taken from Table 1 as given in Eqs. (6-9). Estimated HQ values were used to calculate the hazard index (HI) values for interactions of (i) ABs with metal ions, (ii) ABs with NPs, and (iii) metal ions with NPs in the human digestive system as mentioned in Eqs.
(10-12). Limited information is available in published literature on how ABs interact with metal ions or NPs. Urbaniak et al., (2007) used "k" values to provide information on the strength of interaction between fluoroquinolones and metals and observed strong interaction between them. Few studies reported synergistic effects for the interactions of ABs with NPs (Abo-Shama et al., 2020) or metal ions (Nazari et al., 2012); however, antagonistic effects do occur as well. Turel (2002) in his study showed that the activity of fluoroquinolones reduces in the presence of metal ions. Complexation with metal ions is one of the primary reasons for the reduced activity of FQs (Seedher and Agarwal, 2010), as complexation modifies their solubility and binding capacity (Djurdjevic et al., 2007). This study assumed that synergistic effects occur for the interaction of ABs with metal ions or NPs. For cases (i) and (

Risk management
Maximum allowable concentration (C max ) can be defined as the concentration beyond which no adverse effects or risk exposure can occur. C max specifies the upper limit values of substance under study and can aid regulatory agencies for managing the risk. To calculate the C max values, the HQ values in Eqs. (6-9) were set as 1, and the concentration was calculated.

Estimation of NPs and ABs loading in the digestive media
The results revealed that Fe 3 O 4 NPs showed high bioaccumulation in the digestive system with a value of 2.58 × 10 −3 mg/L. The comparative analysis indicated that Fe 3 O 4 NPs have the highest accumulation in the digestive media which was followed by ZnO NPs (5.77 × 10 −8 mg/L), CuO NPs (2.40 × 10 −9 mg/L), and TiO 2 NPs (7.19 × 10 −10 mg/L). Overall, concentration of Fe 3 O 4 NPs was found to be highest amongst all the NPs considered for this study. High accumulation of NPs like ZnO and CuO in the digestive system can be related to their size values. The larger the size of NPs, the higher is the accumulation potential in the human digestive system (Bergin and Witzmann, 2013). The observed sequence of NPs in the human digestive system is similar to those observed by Parsai and Kumar (2020). Similar to NPs, the concentration of ABs in the human digestive media (C AB_DM ) after NPs dissolution was also calculated to determine the accumulation potential of ABs in the human digestive system. The concentration of levofloxacin cannot be determined due to the lack of adsorption isotherm data. Risk can be calculated in the future once data is available. Amongst the ABs considered, OFX (1.8 × 10 −3 mg/L) showed highest accumulation potential in the human digestive system next to CIP (3.87 × 10 −7 mg/L) and NOR (4.44 × 10 −7 mg/L). Wingender et al., (1985) reported rapid absorption of CIP in the upper GI-tract; however, comprehensive information on the absorption of CIP or any other FQ in different parts of the human GI-tract does not exist (Sharma et al., 2017). Research studies suggest (10) HI = HQ antibiotics + HQ metal ions (11) HI = HQ antibiotics + HQ NPs (12) HI = HQ metal ions + HQ NPs that accumulation of FQs antimicrobial agents is reduced by lowering the pH and, under some conditions, by divalent cations as well (Vergalli et al., 2020).

Risk estimation of metal ions released in the digestive media after dissolution of NPs
The results revealed that the HQ values for all types of metal ions released in the human digestive system after NP dissolution were observed to be less than 1 under the conditions assumed in this study. This indicated that the metal ions released from the NPs in the human digestive system do not show any significant health risks to human health [HQ values ranged from 1.44 × 10 −11 (for Ti 2+ ions) to 2.21 × 10 −4 (for Fe 3+ ions)].

Estimation of risks due to the revised concentration of NP in the digestive media after dissolution of NPs
HQ values calculated for the revised concentration of NPs in the human digestive system were in the sequence of (low to high): ZnO NPs, 3.43 × 10 −8 ; TiO 2 NPs, 3.19 × 10 −7 ; CuO NPs, 2.28 × 10 −5 ; Fe 3 O 4 NPs, 2.85 × 10 −3 ( Table 2). The results showed that the estimated HQ values for the revised concentration of NPs were smaller than 1, indicating no significant risks to human health. Amongst the NPs considered, Fe 3 O 4 NPs showed comparatively high HQ values than other NPs. It is important here to mention that till the time of writing, no such studies (as per the author's best knowledge) have been reported in literature to determine risks for the released concentration of metal ions as well as NPs in the human digestive system. It is also not known how the released metal ions behave after their dissolution from NPs in the digestive media. Moreover, no guidelines or recommendations are available on the use of nanomaterials as a coating agent for drug delivery. In spite of being used in numerous biomedical and other industrial uses, the safety, toxicity, and its interaction with and within the biological systems are still unclear (Snyder-Talkington et al., 2012;Zhong et al., 2017).

Risk of revised concentration of ABs in the digestive system after release from NPs and absorption in GI-tract
HQ values estimated using the revised concentration of ABs in the human digestive system ranged from 1.90 × 10 −12 (NOR) to 3.37 × 10 −8 (OFX), an indication of no health risks. Amongst all, it was observed that CIP concentration may pose risks to human health due to low values. Overall, the results of risk evaluations disclosed that there exist no risks to human health for the consumption of ABs-NPs complexes through inadvertent ingestion of contaminated water under the conditions assumed in this study and for the scenarios considered. Although the estimated risk values were below the acceptable risk level, the effect and amount of NPs used as a coating material for ABs need to be regulated prior to human use.

No absorption
The term "No absorption" implies that under the hypothetical scenario considered, no adsorption of ABs on NPs takes place. In this case, the calculated PNEC values observed to be (high to low): NOR (233.80 µg/L) > OFX (53.44 µg/L) > CIP (26.72 µg/L) > LEVO (2.505 µg/L). HQ values ranged from 1.68 × 10 −3 (NOR) to 4.3 × 10 −2 (LEVO) and are observed to be less than the acceptable risk level (HQ < 1). Therefore, these ABs do not pose any significant risks to human health. The results were found to be similar to those reported by the authors (Kumari and Kumar, 2020a) due to sulfamethoxazole, ampillicin, and amoxicillin. Even though the presence of ABs in the water environment poses insignificant risks to human health, the monitoring and risk management of these substances are required in order to protect human beings from their detrimental effects, if any.

100% absorption
The term "100% absorption" implies that under the hypothetical scenario, total adsorption of ABs by NPs occurs. Figure 3 shows the HQ values of individual ABs for different NPs. As can been seen, the HQ values for all the ABs were observed to be less than 1, the acceptable risk, indicating no significant risks to human health. Amongst the NPs studied, high risk values were observed for ZnO NPs compared to others. Similar to the other cases, here also, absorption of ZnO NPs in the human digestive system was found to be the highest. Parsai and Kumar (2020) found high HQ values for ZnO NPs through fish consumption exposure. The toxicity of ZnO NPs might be due to their solubility. It is reported that dissolution of ZnO NPs takes place in the extracellular region, which in turn increases the level of intracellular Zn 2+ .However, the mechanism behind the increased level of intracellular Zn 2+ ions and dissolution of ZnO NPs in the medium is still unclear (Pandurangan and Kim, 2015). This study estimated risks due to the exposure of single type of nanoparticle at-a-time. The chance of occurrence of more than one type of NPs or ABs was not considered due to lack of information on the BAF values of mixture of NPs, dissolution and absorption rate of mixture of ABs in the GI-tract, and dissolution of NPs to their corresponding ions in the digestive system and RfD values. The present study showed an example of determining risk estimates in digestive system after including all the important parameters and highlighted its significance for human health. This type of risk assessment studies has not been conducted, and this is the first attempt to predict risk of ABs and NPs together within the human digestive system. The developed framework will help in assessing risk effects of the exposure of NP coated ABs to human in the digestive system. Though, few studies have estimated risks due the exposure of NPs (Pizzol et al., 2019;Yang et al., 2017) and ABs (Kumari and Kumar, 2020a;2020b) but none of them has considered the effect of human digestive media on the fate of NPs and ABs including the RfD values of NPs. The proposed framework included the effects of dissolution of NPs to their corresponding ions in digestive system and estimated the revised concentration of ABs in the digestive media so as to provide more realistic values of both NPs and ABs in the human digestive system. The suggested framework provides a step-wise approach to determine the risks exposure effects of NPs and ABs-complexes in the human digestive system.

Hazard interaction (HI)
Under the conditions assumed in this study, the ABs-NPs complexes after oral ingestion undergo absorption in the GItract and dissolved to produce different fractions. Therefore, it is essential to calculate the risk of these released components in the human digestive system for getting realistic risk estimates in digestive media. For case (i), HI interaction of ABs with metal ions, the results revealed that the interactions of ABs with metal ions in the human digestive system do not pose any risks to human health (HI int for ABs-metal ion pairs < 1) for all the conditions assumed in this study. In case (ii), HI interaction of ABs with NPs, the calculated HI int values were also observed to be less than 1, indicating no risks. For case (iii), HI interaction of ABs with metal ions, similar to the results obtained for the above two cases, here also the interaction results were smaller than 1, indicating no significant health risks to human health. Overall, it was observed that no significant health was observed for the three mixture pairs as mentioned above. However, guideline values need to be developed so as to regulate the amount and  Table S5 shows information about the results obtained for all the cases and combinations mentioned above. Table 3 provides the calculated maximum allowable concentration for ABs, metal ions, and NPs. C max values of metal ions at which no risks occur ranged from (high to low) 6.68 × 10 5 mg/L (for Ti 2+ ions) to 6.27 mg/L (for Zn 2+ ions). The estimated C max values of NPs were observed to range from 0.658 mg/L (ZnO NPs) to 6.68 × 10 5 mg/L (TiO 2 NPs). Amongst the NPs studied, ZnO NPs showed highest risk, and therefore, the C max values for ZnO NPs were found to be lowest followed by Fe 3 O 4 , CuO, and TiO 2 NPs, respectively. If we compare the C max values of metal ions with that of NPs, it can be seen that the Zn 2+ ions (0.658 mg/L) showed comparatively high C max values than ZnO NPs (6.27 mg/L). Previous studies also reported high C max values for ZnO NPs in water bodies for the inadvertent ingestion of NPs through fish consumption exposure (Parsai and Kumar, 2020). Similar C max values for Ti 2+ ions and TiO 2 NPs, and Fe 3+ ions and Fe 3 O 4 NPs as can be seen from Table 3 is due to the use of similar RfD values for metal ions and NPs (due to unavailability of RfD values for these NPs) during risk estimation. Therefore, there is a need for conducting in vivo and in vitro studies for determining RfD values of NPs to get accurate risk estimates. The observed results demonstrated that stern actions and control measures must be taken to reduce the risk exposure effects of metal ions and NPs. The C max values of ABs were also calculated to determine their allowable concentration in water systems. The study observed that C max values of ABs beyond which no risk effects can occur were observed to be 2.5 µg/L (LEVO), 53.44 µg/L (OFX), 26.72 µg/L (CIP), and 233.80 µg/L (NOR). On the basis of C max values, it can be said that LEVO poses maximum risk to children whereas NOR shows minimum risk. Different C max values of ABs have been reported by researchers (Lubasch et al., 2000;Owens et al., 1997) which might be related to the administered dose of ABs and age of the population considered. The results obtained in this study can be used by the regulatory bodies like USEPA, Organisation for Economic Co-operation and Development (OECD), and WHO for setting up the guidelines values for metal ions, NPs, and ABs in water.

Effect of assumptions used on risk values
Due to lack of information on parameters used to determine risk estimates due to inadvertent exposure of NPs and ABs from natural waters, assumptions were made to fill the data gaps. Firstly, to calculate risk estimates due to the interaction of NPs with ABs, this study assumed the NP concentration to be 1 mg (NPs) per mg of ABs (data taken from literature) to determine the concentration of ABs adsorbed on NPs (mg/L). Under this assumption, it was observed that the estimated risk values for all the conditions studied do not pose any concerns to human health. However, risk estimate might vary as it directly depends on the concentration of substances used for the study (Kumari and Gupta, 2018). Secondly, due to unavailability of RfD values for NPs considered in this study (as mentioned in the exposure assessment section), the RfD values of metal ions were taken (assuming that the values are equivalent and similar to that of NPs) to estimate the risk values. At the assumed RfD values, no risk was observed for human health for the conditions studied. The observed risk estimates might show different results, if estimated using accurate RfD values of NPs. Thirdly, the transformation products and actions of NPs within the human digestive system are hard to anticipate. Fourthly, different values of concentrations of ABs are available. Different values of the selected ABs have been reported by Wang et al., (2018) and Hanna et al., (2018) which create a dilemma on which values to take for determining risk estimates, and thereby creating uncertainty in the overall process. The fifth assumption relates to the value of B ij . The study assumed the B ij values as 1 due to lack of information on the interaction of ABs with NPs or metal ions. The risk estimates presented in this study indicate the point estimate values and can vary depending on variability of these parameters, creating uncertainty in risk estimates. Uncertainty analysis using Monte Carlo simulations needs to be performed to overcome these issues in risk estimates. Moreover, it is essential to recognise those parameters which add high variability in HQ estimation so that efforts could be taken for reducing their variability in risk estimation process.

Summary and conclusions
The major findings of the study are presented below: • The present study proposed a framework for assessing health risk exposure effects caused due to the interaction of NPs and ABs during oral administration. The developed framework was applied to a hypothetical scenario where environmentally occurring concentration of NPs (Fe 3 O 4 , ZnO, CuO, and TiO 2 ) and ABs (LEVO, OFX, CIP, and NOR) was taken for illustrative purpose. • The study estimated the loading of ABs and NPs in the human digestive system after their release. Amongst the NPs, Fe 3 O 4 NPs (2.58 × 10 −3 mg/L) presented maximum accumulation in the digestive media whereas TiO 2 NPs presented the minimum (7.19 × 10 −10 mg/L). Similarly, for ABs, OFX has the highest accumulation rate in the human digestive system (1.8 × 10 −3 mg/L). • The risk estimated for two different scenarios showed HQ values less than 1 under the conditions and assumptions made in this study. Therefore, on the basis of results obtained, it can be said that the interaction of two contaminants does not pose any risks to human health during their release and dissolution in the human digestive system. • Mixture toxicity assessment studies by means of HI interaction studies were conducted for three different binary combinations: (i) ABs with metal ions, (ii) ABs with NPs, and (iii) metal ions with NPs. The estimated HI values for all the mixture combinations were observed to be less than 1, the acceptable limit, and therefore these mixture combinations do not pose significant risks to human health. However, more detailed studies on the interactions, ABs with metal ions and NPs, are required (B ij and M ij values in the WoE approach) for accurate risk predictions.
Overall, this study proposed a framework using a six-step risk assessment approach to determine the risk exposure effects of ABs and NPs during inadvertent ingestion of contaminated water and their possible interaction within the human digestive system. This study provides a systematic information on risk assessment involving the fate of NPs and ABs within the human digestive system. Agencies like USEPA and Food and Drug Administration (FDA) have suggested the use of alternative testing strategies and data requirement for NPs (Aschberger et al., 2016); however, a systematic approach in dealing with NPs and ABs does not exist. The outcome of the study now helps in (i) identifying the possible concentration of released compounds (ABs, NPs, metal ions) in the human digestive system and (ii) formulating guideline limits of both ABs and NPs. Studies by Parsai and Kumar (2021) on NPs and Kumar (2020a, 2020b) have developed risk assessment framework for determining risk exposure effects alone and in mixture combination; however, none of the reported studies analysed the interaction between these two contaminants. In this regard, this study provides an approach for understanding the interaction as well as the fate and behaviour of two contaminants. Besides, the maximum allowable concentration values of contaminants derived in the study can be used by the regulatory bodies to regulate the concentration of ABs, NPs, and metal ions in natural water systems. Efforts are required for conducting proper in vitro and in vivo ecotoxicity studies so that better understanding can be made on the fate and behaviour of released fragments of NPs and ABs in the human digestive system.