To evaluate the content characteristics of lead (Pb), cadmium (Cd) and mercury (Hg) in 18 kinds of edible fungi species from Guizhou Province, as well as the impacts of their bioaccessibility on dietary risks.

Physiologically based extraction test (PBET) was employed to simulate gastrointestinal digestion.Concentration data of the 3 kinds of heavy metals before and after extraction were substituted into a health risk assessment model to quantify changes in their content during the digestive process.

The concentration ranges of the 3 kinds of heavy metals, Pb, Cd and Hg, were 0.0116–0.4575, 0.0031–0.7749 and 0–0.0264 mg/kg,respectively. After incorporating the bioaccessibility data, the ranges decreased to 0–0.1942, 0–0.4548 and 0–0.0187 mg/kg, respectively. Meanwhile, the mean values of the estimated daily intake (EDI) were reduced by 39.05%,55.25% and 80.69%, respectively, and the mean values of the target hazard quotient (THQ) and target cancer risk(TCR) were decreased by 39.05%, 55.25% and 77.03%, respectively, with the same reduction rates observed in both adults and children.

The edible mushroom samples in this batch exhibit a certain degree of heavy metal contamination. However, after simulated gastrointestinal digestion, the results of health risk assessment are further reduced. Special attention should be paid to heavy metal contamination in Lentinus edodes, Agaricus blazei,Morchella esculenta and Dictyophora indusiata. Notably, the dietary health risks to children deserve more attention compared to adults.

To develop a Sophora japonica L. rice cake product using Guangxi-specific Sophora japonica L. as the raw material, optimize its production process, extend its shelf life and establish a corresponding shelf life prediction model.

Single-factor experiments combined with response surface methodology were employed to determine the optimal formulation. Subsequently, 2.5 g/kg calcium propionate was added as a preservative, and the products were vacuum-packaged and stored at 20 °C and 6 °C, respectively. Periodic measurements were conducted for indicators including pH, sensory score, acid value, moisture content and total viable count. A shelf life model was constructed based on zero-order reaction kinetics and the Arrhenius equation,and comprehensive quality evaluation was performed using the fuzzy weight method.

The optimal formulation for Sophora japonica L. rice cake was determined as follows: Rice-to-flour ratio of 2.5:1 (m:m), Sophora japonica L. addition of 1.50%, sugar content of 15.0%, yeast dosage of 0.75% and fermentation time of 50 min. Under storage conditions of 20 °C and 6 °C, the prediction errors of the shelf life model were both below 10% except for the acid value, indicating high accuracy of the model.

This study successfully optimize the production process of Sophora japonica L. rice cake and establish a reliable shelf life prediction model, providing a theoretical basis for its practical production and storage.

To analyze the contamination status and influencing factors of Burkholderia gladioli(Pseudomonas cocovenenans subsp. farino fermentans) and bongkrekic acid in edible fungus in Guangzhou City.

Samples of 480 pieces of dried black fungus and fresh silver fungus were collected from marketplaces,supermarkets, stores, online stores and catering units. Accordance with GB 4789.29—2020 National food safety standard-Food microbiology inspection-Inspection of Burkholderia gladioli (Pseudomonas cocovenenans subsp.farino fermentans), and the strain identification was carried out. The liquid chromatography-mass spectrometry method was used to detect the bongkrekic acid in samples.

It showed that no bongkrekic acid was detected in 480 samples. The detection rate of Burkholderia gladioli was 16.88% (81/480). There was no statistically significant difference in the detection rate of Burkholderia gladioli among food categories, sampling sites, packaging types and materials (P>0.05). The detection rates of Burkholderia gladioli were different in sample origin, sampling quarter and the detection of mold, which were statistically significant (P<0.05). Multivariate logistic regression found that samples from the first quarter (March) had a higher risk of contamination by Burkholderia gladioli than those from the other 3 quarters. The toxicity test identified 5 strains as Pseudomonas cocovenenans subsp. farino fermentans, from 4 samples of wet black fungus and 1 sample of fresh silver ear, all collected in the second quarter.The difference in the sampling quarters of Burkholderia gladioli (Pseudomonas cocovenenans subsp. farino fermentans) was statistically significant (P<0.05).

In the market circulation process, Burkholderia gladioli is prone to grow in foods such as wet black fungus and fresh silver ear fungus, and there is a risk of producing bongkrekic acid. Therefore, it is necessary to strengthen the monitoring of this pathogenic bacterium(Pseudomonas cocovenenans subsp. farino fermentans) in key foods (with particular attention paid to the second quarter during the market circulation process).

To investigate the current contamination status of veterinary drug residues in eggs sold in Guangzhou and assess the dietary exposure risks of different age groups.

From 2017 to 2025, a total of 642 egg samples were randomly collected from farmers’ markets, conventional supermarkets, fresh food supermarkets and online stores in a total of 11 administrative districts of Guangzhou City. Liquid chromatography-tandem mass spectrometry was used to detect 23 kinds of veterinary drugs residues in 6 categories.Based on non-parametric probabilistic assessment method and combined with local residents’ egg consumption data,the dietary exposure risk was assessed using the hazard quotient (HQ) method.

The overall detection rate of veterinary drug residues in eggs sold in Guangzhou was 24.92% (160/642), with an overall exceedance rate of 10.28% (66/642). The top 3 veterinary drugs in terms of the detection rates were doxycycline (8.89%, 56/630),sodium pentachlorophenolate (5.72%, 27/472) and oxytetracycline (5.56%, 35/630). The top 3 veterinary drugs in terms of the exceedance rates were sodium pentachlorophenolate (5.72%, 27/472), metronidazole (2.18%, 14/642)and florfenicol (based on the sum of florfenicol and florfenicol amine, 1.79%, 10/560). Additionally, simultaneous detection and exceedance of 2 or more veterinary drug residues were found in 6.07% (39/642) and 1.87% (12/642) of the samples, respectively. The overall exceedance rate in eggs from farmers’ markets (11.24%, 19/169) was significantly higher than that from online stores (2.94%, 4/136) (P<0.05). The overall exceedance rate in bulk eggs(11.79%, 50/424) was significantly higher than that in pre-packaged eggs (6.90%, 14/203) (P=0.001). The dietary exposure assessment results showed that the HQ values for all detected veterinary drugs were far below 1, indicating that the risks were within an acceptable range. Notably, children aged 3–6 years exhibited relatively higher exposure levels and health risks, which required special attention.

The overall risk of veterinary drug residues in eggs sold in Guangzhou is controllable and the health risk of dietary exposure remains at an acceptable level. It is recommended that regulatory authorities prioritize the monitoring of veterinary drug residues in eggs sold in farmers’ markets and bulk eggs,while simultaneously strengthening education and awareness campaigns regarding the standardized use of veterinary drugs in poultry farming.

To explore the content of multiple elements in Ziziphus jujuba leaf tea and evaluate its nutritional characteristics and heavy metal contamination risk.

Commercially available Ziziphus jujuba leaf tea samples from 7 major production areas were subjected to microwave digestion pretreatment. The content of 15 kinds of elements (K, Ca, Fe, Na, Mn, Sr, Zn, Cu, Cr, Ni, Pb, As, Se, Cd, Hg) were determined by inductively coupled plasma atomic emission spectrometry (ICP-AES), inductively coupled plasma mass spectrometry (ICP-MS)and atomic fluorescence spectrometry (AFS). Pollution assessment and health risk evaluation for 6 kinds of heavy metals (Cu, Cr, Pb, As, Cd, Hg) were conducted using the single factor pollution index method, the Nemerow multi-factor composite index method and the target hazard quotient method.

The average content order of the 15 kinds of elements in Ziziphus jujuba leaf tea was K>Ca>Fe>Na>Mn>Sr>Zn>Cu>Cr>Ni>Pb>As>Se>Cd>Hg.The content of K, Ca and Fe were relatively high, respectively, with these 3 kinds of elements accounting for 97.9% of the total measured element content. Furthermore, the element content in samples from different production areas generally showed regional differences (P<0.05). The content of the 6 kinds of heavy metals did not exceed the limits specified in GB 2762—2022 National food safety standard-Maximum levels of pollutants in foods. The single factor pollution indexes were all below 0.6, the Nemerow comprehensive pollution indexes were all below 0.7, and the target hazard quotients and total hazard index were all less than 1.

Under the evaluation model of this study, the heavy metal pollution level in commercially available Ziziphus jujuba leaf tea is low, and the associated health risks to humans are within an acceptable range.

To assess the safety of a taurine-zinc-selenium fortified walnut milk.

The fortified walnut milk was lyophilized into a powder prior to toxicological evaluation. Its safety was assessed via an acute oral toxicity test and a 28 d repeated-dose oral toxicity study. For the acute toxicity test, specific pathogen free(SPF) Kunming mice (n=20, half male and half female) were administered the test substance at the maximum feasible dose, which was determined by its maximum solubility and gavage volume. The dose was administered in 2 equal portions 6 h apart, followed by a 14 d observation period for clinical signs and mortality. For the 28-day study, a total of 80 SPF Sprague-Dawley rats were randomly allocated to 4 groups and the test substance was administered by gavage at doses of 0 (control), 1.8 (low), 3.6 (medium) or 7.2 g/(kg·bw)/day (high). These dosages corresponded to 4, 8, and 16 times the recommended human intake, respectively. Throughout the study, monitored clinical manifestations, body weight, food consumption and food utilization efficiency. At termination, blood and urine samples were collected for hematological, serum biochemical and urinalysis assessments. Major organs were weighed, and histopathological examinations were performed.

In the acute oral toxicity test, no mortality or treatment-related clinical signs of toxicity were observed. Gross necropsy revealed no organ abnormalities. The median lethal dose (LD₅₀) was therefore determined to be greater than 18.0 g/(kg·bw), classifying the substance as practically non-toxic. In the 28 d oral toxicity study, all animals remained in good health, and there were no statistically significant differences in body weight gain, food consumption, food utilization efficiency,hematological, serum biochemical parameters and organ coefficient between the treatment groups and the control group (P>0.05). All urinalysis parameters were unremarkable. No treatment-related histopathological lesions or cellular damage were found in any major organs.

The taurine-zinc-selenium fortified walnut milk demonstrates a high degree of safety under the conditions of this study. It does not induce acute or subacute toxicity at doses up to 7.2 g/(kg·bw)/day in rats. Based on these findings, the no-observed-adverse-effect level (NOAEL) is established at 7.2 g/(kg·bw)/day, the highest dose tested.

To construct a Bayesian network risk assessment model to systematically quantify the risks of tetracycline, quinolone, β₂-receptor agonist and sulfonamide drug residues.

The risk value was evaluated by expert scoring, and the initial model was established by using the information entropy weight method to obtain the prior probability. The Bayesian network warning model was established by importing the monitoring data through the expectation maximization algorithm (EM) of Bayesian network, and the risk prediction of the risk factors of animal residues was carried out.

The overall risk probability of a province obtained by the Bayesian network model was (within the maximum detection limit): Low risk 74%, medium risk 16% and high risk 10%. The highest sensitivity of tetracycline was 0.37, which was the key risk factor; the risk of quinolones and sulfonamides was in good condition; as a banned drug, the risk probability of β₂-receptor agonist did not reach 100% low risk, and supervision needed to be strengthened.

The constructed Bayesian network model has risk assessment and warning, and many kinds of drugs can be classified and supervised, which provides technical support for improving the accuracy and timeliness of veterinary drug residue risk management.