Environmental and seasonal drivers of microbiological contamination in street-vended foods

Abstract

Street-vended foods are an integral part of urban food systems, providing affordable meals but posing potential public health risks when hygiene and temperature controls are inadequate. This study assessed the microbiological safety of 248 street-vended food samples collected from a touristic area and two institutional neighborhoods in Lebanon during the winter and summer seasons. Samples were analyzed for coliforms, Escherichia coli, Staphylococcus aureus, Salmonella spp., and Listeria monocytogenes using internationally recognized ISO culture-based methods. Coliforms were detected in 42% of samples at acceptable or unsatisfactory levels, while E. coli and S. aureus were found in 6.85% and 4.84%, respectively. Salmonella spp. and L. monocytogenes were not detected. Contamination was more frequent during warmer periods and in densely populated areas, reflecting the influence of environmental factors and crowding on food safety. Comparative analysis with other Mediterranean studies revealed similar contamination profiles, indicating shared regional challenges related to hygiene practices, water quality, and temperature control. The results support a predictive risk modeling framework linking contamination likelihood to crowding, season, and hygiene determinants within informal urban food systems. This evidence contributes to the global One Health agenda by informing risk-based inspection models and cross-sectoral strategies for safe, resilient street food economies.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-32430-y.

Introduction

Street foods are ready-to-eat foods and beverages prepared and/or sold by vendors or hawkers in public places, such as streets and other similar areas^1,2. Nowadays, millions of people are estimated to consume street food in some form every day^3,4. Compared to restaurant foods, street foods are low-priced meals, which makes them easily accessible and affordable for the population in many developing countries³. Street foods are also an attraction for tourists, as they are reasonably priced and provide a convenient substitute for home-cooked meals during their vacation^3,5.

Street foods are recognized as hazardous products posing significant public health risks², with microbiological contamination being the most critical health hazard³. This results from many factors, such as a lack of basic infrastructure, vendors’ poor knowledge of fundamental food safety measures, and insufficient public awareness of the possible hazards from street foods². Furthermore, most countries lack governmental oversight or control over the vast number of street food vending businesses⁶, making it extremely challenging to manage them, especially in overcrowded streets and public spaces³. It is essential to acknowledge that many individuals from low-income households rely heavily on their street food vending businesses to generate their income⁷. To further support this statement, many people in Lebanon had to find alternatives after losing their employment, particularly in the wake of the recent severe economic crisis that hit the country⁸. This was demonstrated by the opening of several street food outlets, especially considering that starting a street food business is way less expensive than opening a restaurant⁹.

Despite their growing role in urban nutrition worldwide, few studies have quantified environmental and behavioral determinants of microbial contamination within informal food systems. To our knowledge, no study targeted kiosks and snack bars in Lebanon. This said, there is a need for research on street food safety. Accordingly, this study targeted two cities in Lebanon, Beirut and Batroun, to compare the safety status of the food sold in kiosks and snack bars in crowded seasons. In Beirut, the snack bars around universities were selected since those areas are usually the most visited by students and workers. In Batroun, the kiosks present during the Christmas festival and the summer season were selected when the highest number of local and international tourists were expected to visit the city. The study aims to identify the health threats associated with eating street food and compare the different circumstances that induce food contamination, considering the frequency of sales throughout the year and the setup of the places. Three objectives were targeted: (i) to quantify the microbiological contamination levels of street-vended foods across different urban and seasonal contexts, (ii) to assess the influence of environmental and operational factors, including seasonality, crowding intensity, and vending setup, on microbial contamination, and (iii) to generate evidence-based recommendations to inform food-safety regulations and guide policymakers in strengthening the safety and resilience of the street-food sector.

Materials and methods

Sampling locations

Batroun, a coastal city in northern Lebanon, was chosen as the first destination for this study. This city has recently become a thriving tourist destination. It is situated by the beach, attracting both local and international tourists, especially during the hot summer months. In this city, all the kiosks present were targeted for testing. Restaurants were not included in the study.

The second destination is Beirut. The study targeted two university neighborhoods as they are the most crowded: the Bliss area near the American University of Beirut (AUB), considered crowded all year round due to its proximity to many hospitals in addition to the university, and the Tarik Al Jadideh neighborhood near the Beirut Arab University (BAU), mostly crowded during the university teaching semesters. In those two areas, 15 snack bars were targeted, located within a maximum 10-minute walk from the universities. The snack bars were considered places with at least one missing wall, exposing the preparation area to the street. Restaurants that were closed were not included in this study.

Food samples analysis

Sampling

The study was conducted during the winter and summer seasons between December 2023 and June 2025 (Table 1). In Batroun, the samples were collected during the Christmas season of 2023 (December 2023) and the summer season of 2024 (July 2024). Sixty-eight food samples, 2 to 3 samples from each vendor, were purchased during both seasons (24 during Christmas and 44 during the summer) from 33 street food selling points, covering all kiosks during the two seasons (8 kiosks in Winter and 15 in Summer). In Beirut, 90 samples were collected from 30 food vending facilities within a 10-minute walk from each targeted area, AUB and BAU, during winter (January through March 2024) and summer (August 2024 for the Bliss area and June 2025 for the Tarik Al Jadideh area). Three samples were chosen from each vending facility. The test was done in triplicate.

Table 1.

Sample collection for the targeted areas.

Area		Code	Season	Date	Vending facilities	Number of samples
North Lebanon	Batroun	NLB	Winter	December 2023	8	24
			Summer	July 2024	15	44
Beirut	Bliss Street	BBS	Winter	January to February 2024	15	45
			Summer	June 2024	15	45
	Tarik el Jadideh	BTJ	Winter	February to March 2024	15	45
			Summer	June 2025	15	45

Samples were purchased, stored under controlled conditions at 4˚C, and transported to the Food Safety Facility at the American University of Beirut. The food samples were grouped into subgroups as proposed by Barreira et al. (2024). The classification was mainly based on the type of preparation (fully cooked, mixed fully cooked/pasteurized, or raw), the presence of raw components, and whether the food was handled after heat treatment (Table 2). Supplementary Table S1 details the street food samples collected during the winter and summer seasons.

Table 2.

Food samples’ classification into groups and subgroups.

Group	Sub-group	Samples	Description
Group 1	1 A	83	Fully cooked food not handled after heat treatment
	1B	11	Fully cooked foods handled after heat treatment not served hot
	1 C	9	Fully cooked foods added with pasteurized components that were stored under refrigeration during shelf-life
Group 2	2 A	16	Mixed fully cooked/pasteurized foods added with raw fruits, with or without sauces
	2B	109	Mixed fully cooked/ pasteurized foods added with raw vegetables, and may include raw fruits
Group 3	3 A	20	Raw fruits and/or raw vegetables, cut or prepared for consumption on the same day, with or without sauces, added, or not, with small quantities of fully cooked food

Group 1 food samples (n  = 103) consisted mainly of snacks or appetizers (e.g. cheese rolls), pastry products (e.g. pancakes), baked dough (e.g. Kaake and Man’ouche) and pasta meals.Group 2 food samples (n = 125) included mainly sandwiches containing fresh vegetables (e.g. hamburgers, shawarma and burritos all containing vegetables), and desserts with dairy products and fruit toppings (e.g. milkshakes).Group 3 food samples (n = 20) included fruits and juices with, and fresh salads.

Microbiological analysis

Twenty-five grams of each sample were suspended in 225 mL of sterile Buffered Peptone Water (BPW) (Bio-Rad, United States) and homogenized using a stomacher for 2 min, resulting in a 10^ (− 1) dilution. Serial dilutions were then prepared. One hundred microliters of each dilution were inoculated onto RAPID’E.coli agar plates (Bio-Rad, United States) for the detection and enumeration of E. coli and coliforms (detection based on the colony’s color on the agar), and Baird Parker agar plates for the detection and enumeration of Staphylococcus aureus. Pure colonies were characterized based on morphological differences after incubation at 37 °C for 24 h. Confirmation of the isolates was conducted using API Staph and API 20E (Biomérieux, France) in addition to Latex confirmation tests for S. aureus (Bio-Rad, USA).

To test for the presence of Salmonella sp., 25 g of each sample was mixed with 225 mL of BPW (Bio-Rad, United States), homogenized using a stomacher, and incubated at 41.5 °C for 24 h. One mL of the BPW-mixture was then aseptically transferred into 9 mL Rappaport Vassiliadis Soybean meal broth RVS (Bio-Rad, United States) enrichment broth, homogenized, and incubated at 41.5 °C for 24 h. The next day, 100 µL of the RVS mixture was aseptically streaked onto RAPID Salmonella agar plates (Bio-Rad, United States) and incubated at 37 °C for 24 h. Confirmation of the isolates was conducted using the API 20E system and the Latex confirmation test for Salmonella (Bio-Rad, USA).

For Listeria monocytogenes, 25 g of each sample was mixed with 225 mL of half-Frazer broth (Scharlau, Spain) and incubated at 30 °C for 24 h. The mixture was then aseptically streaked onto RAPID’Listeria (Bio-Rad, United States) and incubated for 24 h at 37 °C. Confirmation of the isolates was conducted using API LISTERIA (Biomérieux, France).

Interpretation of the microbiological results

Due to the lack of local Lebanese standards, the microbial quality of the food samples was compared to two standards: (1) Australian and New Zealand guidelines for Microbiological Examination of Ready-to-Eat Food¹¹ and (2) Chinese Microbiological Guidelines for Food - For ready-to-eat food in general and specific food items¹². The criteria for microbiological quality classifications are compiled in Supplementary Table S2.

Statistical analysis

Data were analyzed using IBM SPSS Statistics version 28.0 (IBM Corp., Armonk, NY, USA) and verified in R (version 4.3.2). Microbiological results were expressed as categorical variables (good, acceptable, or unsatisfactory) and further dichotomized for statistical analysis, where good results were coded as not contaminated (0) and both acceptable and unsatisfactory results were coded as contaminated (1). Associations between contamination status and categorical factors (sampling site and season) were evaluated using the Pearson Chi-square (χ²) test of independence. Statistical significance was set at p < 0.05.

Descriptive statistics (frequencies and percentages) were used to summarize the distribution of contaminated samples across sites and seasons. Results were presented as χ² test statistics with corresponding degrees of freedom (df) and p-values.

To quantify environmental determinants, we computed odds ratios (OR) and risk ratios (RR) with 95% confidence intervals from 2 × 2 contingency tables (season × contamination and organism-specific contamination), and pairwise site effects using BTJ as reference. ORs were estimated with Woolf’s log-method (continuity correction applied for zeros); RRs used Wald CIs on the log scale.

Results

The microbiological quality of street food samples purchased from two cities in Lebanon during the winter and summer seasons was evaluated. This involved detecting key pathogens and enumerating indicator microorganisms such as coliforms. When compared to the two microbiological guidelines for Ready-to-Eat Foods, the samples collected from the two cities during the two tested seasons showed varying levels of acceptability, depending on the microorganisms detected.

Evaluation of the general food hygiene

The indicator microorganisms (coliforms) were detected at acceptable levels in 21.77% of the total samples and at unsatisfactory levels in 20.56%. Only 57.66% were not considered contaminated with coliforms. This raises the question of applying general Good Hygiene and Good Manufacturing practices in the outlets. When comparing the three sampling sites, it was noticed that the level of unsatisfactory coliform levels detected at BBS was higher than the two other sites (30% of the BBS samples compared to 17.65% at NLB and 13.33% at BTJ), taking into account that BBS is a year-round busy area (Fig. 1).

Fig. 1.

Coliform contamination detection at the three sampling sites. Total indicates the comparison of the samples across all sampling areas. BBS represents the results at Beirut Bliss Street, BTJ represents the Beirut Tarik el Jadideh area results, and NLB represents the North Lebanon Batroun area results. The results are represented as a percentage of samples in each group: Good, acceptable, or unsatisfactory.

When comparing the sampling seasons, no differences were determined within the three contamination groups. Around 57% of the total samples were not considered contaminated with coliforms during both seasons, 23.13% of the summer samples and 20.18% of the winter samples were within acceptable limits, and 19.4% of the summer samples and 21.93% of the winter samples were considered holding unsatisfactory levels of contamination (Fig. 2). However, differences were observed when comparing the general level of food contamination within the sampling areas. In fact, the summer season harbored more contamination at BBS than in the other two areas. Only 48.89% of the BBS summer samples were not contaminated, 13.33% were within acceptable levels, and 37.78% were considered unsatisfactory. The BBS winter samples were less contaminated, with 53.33% within good limits, 24.44% acceptable, and 22.22% unsatisfactory. The BTJ area was less contaminated, with 66.67% of the summer samples within good limits, compared to 53.33% in winter.

Fig. 2.

Coliform contamination per sampling site per season. Total indicates the comparison of the samples across all sampling sites, regardless of the season (summer and winter). BBS represents the results at Beirut Bliss Street during both seasons, BTJ represents the Beirut Tarik el Jadideh site results, and NLB represents the North Lebanon Batroun site results. The results are represented as a percentage of samples in each group: Good, acceptable, or unsatisfactory.

Additionally, 26.67% of the summer samples were acceptable, compared to 22.22% in winter, and only 6.67% were contaminated in summer, compared to 20% in winter. Finally, in the NLB area, more contamination was detected during summer, with only 56.82% within good limits compared to 66.67% in winter. 29.55% of the summer samples and only 8.33% of the winter samples harbored acceptable contamination levels. 13.64% of the summer and 25% of the winter samples were considered contaminated at unsatisfactory levels (Fig. 2).

Finally, when comparing the level of contamination concerning the different food groups, the expected scenario was observed (Fig. 3). Food group 1, which contained fully cooked samples, was less contaminated than food group 2, which consisted of fully cooked samples mixed with fresh fruits and vegetables after cooking. However, all the food group 3 samples were contaminated with coliforms at acceptable or unsatisfactory levels.

Fig. 3.

Coliform contamination per food group. Food Group 1 represents fully cooked products mixed with pasteurized ingredients. Food Group 2 represents the mixed fully cooked/pasteurized products with raw ingredients. Food group 3 represents uncooked food.

Pathogen detection in the food samples

While detecting coliforms in food can indicate the general hygiene practices during preparation, identifying the presence of pathogens is crucial for assessing the health risks associated with consuming contaminated food. Interestingly, neither Salmonella sp. nor L. monocytogenes was detected in any of the food samples collected from the three areas during either season. In addition, very few samples were contaminated with S. aureus and E. coli.

Two winter samples from NLB belonging to food group 1 were contaminated with S. aureus at unsatisfactory levels, and ten samples were contaminated at acceptable levels. Of the ten samples, six were winter samples belonging to food groups 1 (5 BTJ) and 2 (1 NLB), and four BBS summer samples belonging to food groups 2 (75%) and 3 (25%). Regarding the fecal contaminant E. coli, sixteen samples were found to be contaminated at unsatisfactory levels, and one sample was found to be at an acceptable level. From the unsatisfactory group, eight were from winter samples belonging to food groups 1 (3 BTJ), 2 (3 BTJ and 1 BBS), and 3 (1 BBS). In parallel, eight were from summer samples belonging to food groups 1 (1 BTJ, 2 (4 NLB), and 3 (2 BBS and 1 NLB). The sample with an acceptable level of contamination was from a winter sample from BTJ belonging to food group 1 (Supplementary Table 1).

Association between sampling Site, Season, and contamination

A total of 248 food samples were examined to evaluate the effect of sampling site and season on overall microbiological contamination. Of these, 109 samples (43.9%) were classified as contaminated (acceptable or unsatisfactory) and 139 (56.1%) as not contaminated (Table 3).

Table 3.

Association between sampling site, season, and overall contamination.

Factor	Category	Contaminated (n)	Not Contaminated (n)	χ²	df	p-value	p-value
Site	NLB	29	39	1.51	2	0.47	0.47
	BBS	44	46
	BTJ	36	54
Season	Winter	94	20	1.78	1	0.18	0.18
	Summer	100	34
Coliforms	Winter	104	30	0.72	1	0.40	0.40
	Summer	97	37
E. coli	Winter	7	127	0.00	1	1.00	1.00
	Summer	8	126
S. aureus	Winter	8	126	0.79	1	0.38	0.38
	Summer	4	130

Contamination frequencies were 42.6% at the touristic site (NLB), 48.9% at the crowded institutional site (BBS), and 40.0% at the less crowded institutional site (BTJ). Although contamination appeared slightly higher in the densely populated institutional area, the Chi-square test of independence showed that this difference was not statistically significant (χ² = 1.51, df = 2, p = 0.47). Thus, the overall contamination rate did not differ significantly among the three locations.

When comparing seasons, 82.5% of winter and 74.6% of summer samples were contaminated. The Chi-square test revealed no significant association between contamination and season (χ² = 1.78, df = 1, p = 0.18). Despite a numerically higher contamination rate in winter, the variation was not statistically meaningful, suggesting that seasonal temperature differences alone may not explain contamination patterns.

For coliforms, contamination reached 77.6% in winter and 72.4% in summer, with no significant seasonal difference (χ² = 0.72, df = 1, p = 0.40). For E. coli, contamination rates were 5.2% in winter and 6.0% in summer, also showing no association with season (χ² = 0.00, df = 1, p = 1.00). For S. aureus, 6.0% of samples were contaminated in winter and 3.0% in summer, again without statistical significance (χ² = 0.79, df = 1, p = 0.38). None of the samples tested positive for Salmonella spp. or Listeria monocytogenes, confirming the general absence of high-risk pathogens.

These results indicate that street-vended foods in urban Lebanon exhibit contamination patterns comparable to other Mediterranean countries, where overall microbial loads are moderate but attributable primarily to handling and environmental hygiene rather than temperature or geography.

Environmental determinants and risk effects

Overall contamination was higher in winter than in summer, with similar directional patterns for coliforms. No seasonal effect was evident for E. coli, and S. aureus showed a non-significant winter excess. Compared with the less-crowded institutional area (BTJ), contamination risk was higher in the crowded institutional area (BBS) and minimally elevated in the touristic area (NLB). Although confidence intervals include unity, the directional effects consistently implicate crowding/handling rather than season alone as the primary determinant of risk (Table 4).

Table 4.

Association between environmental factors (Season and Site) and Microbiological contamination in Street-Vended foods values represent odds ratios (OR) and risk ratios (RR) with 95% confidence intervals (CI) for contamination by season (Winter vs. Summer) and by site (Institutional and touristic Zones).

Seasonal risk (Winter vs. Summer)	Prev_exposed_%	Prev_ref_%	OR (Woolf)	OR 95% CI	RR	RR 95% CI
Overall contamination	82.5	74.6	1.6	[0.86, 2.97]	1.1	[0.97, 1.26]
Coliforms	77.6	72.4	1.32	[0.76, 2.3]	1.07	[0.93, 1.23]
E. coli	5.2	6	0.87	[0.31, 2.47]	0.88	[0.33, 2.34]
S. aureus	6	3	2.06	[0.61, 7.02]	2	[0.62, 6.48]
Site effect	Prev_site_%	Prev_BTJ_%	OR	OR 95% CI	RR
NLB vs. BTJ	42.6	40	1.12	[0.59, 2.11]	1.07
BBS vs. BTJ	48.9	40	1.43	[0.79, 2.59]	1.22

Discussion

Street food has become increasingly popular in Lebanon, particularly in business and tourist areas, as well as during seasonal events, offering both convenience and a lively atmosphere for consumers. However, alongside its appeal, street food also carries potential risks of foodborne illnesses due to the informal nature of street food vending and poor regulatory oversight¹³. This study assessed the microbiological quality of street foods sold in two cities in Lebanon. The first city, Batroun, is a tourist destination, and the second is Beirut, a business destination. In Beirut, two areas were selected for comparison to assess the impact of crowding on ensuring food safety. Additionally, the testing was conducted over two seasons to assess the impact of environmental temperatures on food contamination.

Across all areas and during both seasons, coliforms’ presence in food indicates contamination resulting from poor hygiene practices or environmental sources, such as soil or water¹⁴. Coliforms usually indicate improper handling, processing, or storage of cooked food, such as chicken or meat burgers, wraps, or sandwiches. Many of the samples belonging to food groups 2 (20.97%) and 3 (65%) showed unsatisfactory levels of coliforms, indicating contamination from washing with contaminated water, inadequate washing, poor storage conditions, or contamination by food handlers during processing. The presence of coliforms in the street food samples was consistent with findings from many other studies. For instance, 51% of the examined sandwiches collected from the streets in Egypt showed several levels of contamination with Enterobacteriaceae¹⁵. Similarly, recent studies conducted in Portugal¹⁰ and Ethiopia¹⁶ have shown the presence of high levels of Enterobacteriaceae and coliforms in various groups of street-vended foods.

The results showed good compliance with GHP among the vendors across both collection seasons since more than 50% of the samples showed no coliform contamination. However, detecting E. coli and S. aureus at unsatisfactory or acceptable levels indicates the need to encourage the food vendors to review or implement GHP. Our findings showed lower contamination levels (6.85%) than the ones presented in a study in Ethiopia, where E. coli was the second most abundant species found in street foods, contaminating almost 20% of the samples¹⁷. In Morocco, a study was done on 224 ready-to-eat samples. The results showed 28% contamination with E. coli¹⁸. Similarly, a study in India found 25.4% E. coli contamination in chutneys, salads, paani puri, and chicken preparations¹⁹. However, in a recent study conducted in a city in North Lebanon, the collected street food samples showed similar ranges of E. coli concentrations as our findings²⁰. In our study, the prevalence of E. coli in food was consistent across both seasons and the three food groups. However, other studies have reported more contamination during the wet-hot season, when vendors lack adequate refrigeration, particularly in food belonging to group 3 ^21–23. The literature reported that the leading reasons for E. coli contamination in street food were fecal contamination of water and produce used during preparation, poor personal hygiene, and possible cross-contamination after heat treatments^13,19,24,25.

Interestingly, the presence of S. aureus, primarily during the winter season, is linked to human contact and poor hygiene practices during food handling and preparation. Unsatisfactory levels of S. aureus suggested that time/temperature abuse likely occurred, allowing the bacteria to grow after improper handling during food preparation^11,24,26,27. The levels of S. aureus detected within these samples were significantly lower (4.84%) than in another recent study in Ethiopia, where S. aureus was the most frequently isolated species in 43.4% of the samples¹⁷. In Namibia, 52% of the samples were contaminated with S. aureus²⁴. Similarly, a recent study in Cameroon, conducted on Ready-To-Eat (RTE) Foods collected from street food vendors, showed that 38.3% of the RTE food samples were contaminated with detectable levels of S. aureus²⁸. Interestingly, a study conducted in North Lebanon²⁰ found that more than half of the collected street food samples were contaminated with S. aureus. Our findings showed that most S. aureus contamination occurred during winter. This result was similar to a study conducted in Italy, which investigated the presence of S. aureus in 3,604 food samples. They found that the bacterial count was the highest during the winter season²⁹.

The absence of Salmonella sp. and L. monocytogenes in our study during both seasons contrasts with many other studies reported in the literature, conducted in Lebanon³⁰, Ethiopia¹⁷, Egypt¹⁵, and West Africa³¹, but was in accordance with other studies done in North Lebanon²⁰ and South Africa³².

Although the differences between sites and seasons were not statistically significant, the numerical variations in contamination levels can be explained by contextual environmental factors. BBS exhibited the highest contamination, particularly in summer, likely due to its consistently dense crowding, narrow streets, heavy traffic, and street-exposed food outlets, as well as the presence of numerous stray cats, which may contribute to environmental contamination. These conditions were reflected in the elevated coliform levels detected in this area. BTJ, by contrast, is less crowded overall, with activity mainly driven by BAU students during the academic year; despite similar street exposure, its environmental load is lower. However, BTJ showed the highest winter E. coli prevalence, which coincides with peak student presence. NLB presented a different pattern, with E. coli more common in summer and S. aureus detected in winter. Foul odors noted during summer sampling suggest inadequate wastewater management, which may explain the seasonal rise in fecal indicators.

To place the Lebanese findings in a regional context, Table 5 presents a detailed comparison with studies from other Mediterranean countries.

Table 5.

Regional comparison of Microbiological contamination in street foods, illustrating how Lebanese findings align with or diverge from studies conducted in other mediterranean countries.

Country / Study	Food Type / Context	Key Microbiological Findings	Comparison With Current Lebanese Study
Portugal 33	Ready-to-eat street foods	• E. coli: 20%• Listeria monocytogenes: 20%• Salmonella & coagulase-positive Staphylococci: absent	Similar absence of Salmonella and L. monocytogenes; higher E. coli prevalence in Portugal.
Portugal10 – (Lisbon)	Street foods in urban Lisbon	• Moderate coliform loads• Contamination linked to poor hygiene and water quality	Aligns with observed Lebanese patterns where coliforms reflect environmental and handling hygiene.
Morocco 18	Street-vended foods in Marrakech	• E. coli: 28%• S. aureus: 22%	Much higher contamination than Lebanon (E. coli 6.85%, S. aureus 4.84%); suggests cleaner practices or faster turnover in Lebanon.
Italy 34	Cooked kebabs in Palermo & Messina	• Occasional E. coli and S. aureus• No Salmonella or L. monocytogenes	Matches Lebanese findings for fully cooked items; pathogens largely absent.
Greece 35	Prepacked sandwiches	• Very low E. coli prevalence• Contamination linked to time–temperature abuse and storage	Supports role of handling/storage practices as determinants rather than season alone.
Spain 36	Ready-to-eat foods	• Coliform contamination associated with poor food-handler hygiene	Consistent with Lebanese findings identifying coliforms as indicators of environmental/handling hygiene.
Lebanon 30	Street foods in Beirut	• Variability in hygiene performance• E. coli and coliforms common indicators	Findings align with the present study’s identification of coliforms and E. coli as key contamination markers.
Interpretation Across Studies	—	• Contamination patterns often linked to handling, water quality, and environmental exposure rather than seasonality	Reflects our study’s lack of significant seasonal association, likely due to predominance of coliforms over temperature-driven pathogens.

These comparisons demonstrate that informal food vending systems across the Mediterranean share recurrent microbial hazards, mainly E. coli and S. aureus, and common vulnerabilities linked to water quality, vendor hygiene, and environmental exposure, rather than inherent differences in climate or geography. The relatively lower contamination rates found in Lebanon reflect potential benefits from rapid food turnover and emerging hygiene awareness among vendors. Yet, they also highlight the need for harmonized regional standards, vendor education, and coordinated surveillance to sustain safe and economically viable street food systems across the Mediterranean.

The modeling outputs generated in this study, particularly the odds ratios and risk patterns across sites, offer a transferable framework for estimating contamination risk under varying urban conditions. By incorporating crowding intensity, season, and site typology as predictors, the results show that environmental exposure and post-processing handling exert a greater influence on contamination than ambient seasonality alone, a finding consistent with Mediterranean evidence linking vendor hygiene and water quality to microbial hazards in informal food networks. Within a One Health perspective, these results point to upstream determinants such as safe water access, waste management, and hand-hygiene infrastructure as critical levers for reducing contamination at the human–environment interface. The effect sizes derived from the model can support probabilistic simulations, predictive surveillance tools, and risk-based inspection planning—such as prioritizing high-crowding zones, strengthening time–temperature control, and implementing targeted seasonal hygiene campaigns. This modeling-informed approach aligns with Codex risk management principles and WHO One Health strategies. It provides a solid foundation for the policy and regulatory actions outlined in the following section.

Recommendations and policy implications

This study assessed the microbiological safety of street-vended foods in Lebanon to evaluate the application of basic food safety practices across different urban settings. Although contamination levels were generally lower than those reported in other Mediterranean studies, the growing popularity of street food underscores the need for proactive risk-based monitoring to prevent foodborne illness and strengthen sector resilience. Building on the modeling insights and the One Health interpretation of the findings, it is evident that reducing microbial hazards requires coordinated actions across multiple levels of the food system: vendors influence hygiene and handling practices; municipalities shape environmental and operational conditions; and national authorities provide the regulatory and infrastructural framework needed to sustain safe informal food markets. Accordingly, the following recommendations are organized across vendor-level, municipality-level, and nation-level interventions to guide targeted improvements and support the development of an evidence-informed, risk-based food safety system.Street food vendors should undergo extensive food safety training certified by the Ministry of Public Health and municipalities. The training should cover personal hygiene, water usage, proper waste disposal, cross-contamination prevention, and temperature control. Conversely, municipalities should provide access to potable water and designated food handling areas, especially in tourist zones and mobile stations. Municipalities should also support vendors in kiosks with refrigerators to maintain the cold chain.

At the vendor level, improving day-to-day hygiene and handling practices is essential to reducing microbial contamination. Vendors should strengthen personal hygiene practices, including regular handwashing, glove use, and wound coverage, to minimize S. aureus transmission. Post-processing handling must be improved by reducing bare-hand contact and ensuring that utensils, cutting surfaces, and preparation areas are kept clean, especially in high-crowding zones. Maintaining strict time–temperature control (keeping cooked foods hot (> 60 °C) and cold foods adequately chilled) is also critical. Access to safe water for washing produce, utensils, and hands should be ensured, with filtered or bottled sources used when municipal water is unreliable. Simple protective measures such as food covers and physical barriers can further reduce environmental contamination, while periodic hygiene coaching, particularly before peak seasons, can support sustained improvements.

Municipal authorities play a central role in shaping the safety of street-food environments. Inspections should be risk-based, with higher frequency in high-crowding locations such as BBS, guided by the risk patterns identified in this study. Standardized hygiene checklists tailored to street-food conditions should be implemented, covering water safety, waste handling, and post-cook contamination risks. Improving wastewater management and street sanitation, especially in areas where foul odors or poor drainage were noted, is essential to reducing environmental contamination. Municipalities should also coordinate seasonal hygiene campaigns, focusing on S. aureus and coliform control in winter and reinforcing temperature control and safe storage during summer. Additional measures include vendor licensing programs with embedded food-safety orientation, partnerships with local laboratories for periodic microbiological screening, and public transparency dashboards that communicate hygiene performance and build consumer trust.

At the national level, there is a need to establish microbiological guidelines explicitly tailored to street-vended foods, adapting Codex RTE standards to the realities of informal food systems. National One Health surveillance frameworks should integrate street-food monitoring, linking food safety with environmental hygiene and public health data. A standardized training and certification system for street vendors, led by the Ministry of Health or the Ministry of Tourism, would help professionalize the sector and improve compliance. Investments in infrastructure, including safe water access, waste management, and upgraded vending stations, should be incorporated into broader urban development strategies. National authorities should also promote data-driven policymaking through predictive risk modeling and multi-season risk assessments, and fund ongoing research on street-food safety, consumer behavior, and environmental determinants to ensure that guidance remains current and evidence-based.

To operationalize a One Health-oriented, risk-based approach to urban food safety, the study proposes a simplified risk-prioritization matrix linking microbiological hazards with corresponding preventive and control actions (Table 6). This framework aligns with SDG 3 on good health and well-being and SDG 11 on sustainable cities, reinforcing the interconnection between food safety, urban planning, and socioeconomic resilience.

Table 6.

One Health-based risk-prioritization matrix for managing Microbiological hazards in informal urban food systems.

Hazard (severity)	Likelihood	Priority actions
Coliforms (low–moderate severity; hygiene indicator)	Common (≈ 75% across seasons)	Hand-hygiene training; utensil/water sanitation; surface cleaning SOPs
E. coli (moderate severity; fecal indicator)	Low (~ 5–6%; no seasonal effect)	Produce washing water quality; raw-veg segregation; glove use
S. aureus (moderate severity; handler-linked)	Low (3–6%; winter signal, wide CI)	Hair/skin/contact controls; ready-to-eat handling; covering of wounds
Salmonella/Listeria (high severity)	Not detected	Keep HACCP vigilance for poultry/eggs/dairy; rapid corrective action if detected

Hazard severity reflects the pathogenic potential and public-health relevance of each microorganism. Likelihood estimates are derived from this study’s observed prevalence data and relative risk ratios. Priority levels integrate severity and likelihood, with additional weighting for environmental determinants (e.g., crowding and season). Recommended actions are aligned with Codex Alimentarius risk-management principles and focus on vendor training, hygiene infrastructure, and preventive inspection frequency.

The proposed matrix provides a pragmatic framework for integrating microbiological evidence into preventive food control systems. Local authorities can move from reactive enforcement to risk-based, data-driven inspection models by prioritizing hazards based on their health significance and environmental likelihood. Embedding this approach into municipal and national programs through standardized hygiene training, periodic laboratory verification, and transparent reporting can enhance consumer trust and vendor compliance. This model illustrates how One Health principles can be operationalized at the street-food level, linking microbiological surveillance, urban planning, and public-health policy. Such alignment strengthens national food-safety governance and contributes to regional efforts toward harmonized microbiological criteria and resilient urban food systems across the Mediterranean.

Conclusions

This study evaluated the microbiological safety of street-vended foods sold across three Lebanese districts, representing both touristic and institutional zones, during winter and summer. Street food in Lebanon remains a cultural hallmark and a vital source of livelihood. Overall, the results indicated satisfactory microbiological quality, with no detection of major pathogens such as Salmonella spp. or Listeria monocytogenes. Nevertheless, the detection of E. coli, coliforms, and S. aureus in several samples at levels exceeding acceptable limits reflects lapses in hygiene and post-processing handling that varied across sites and seasons.

These findings underscore the urgent need for national microbiological criteria and operational guidelines designed explicitly for street food vendors. Beyond laboratory monitoring, sustained improvements in food safety require integrating hygiene education, vendor training, and risk-based inspection systems into municipal and national food control frameworks. Local authorities should adopt preventive and participatory approaches supported by regular microbiological surveillance and transparent reporting mechanisms.

From a broader policy perspective, strengthening the microbiological safety of street food in Lebanon extends beyond a public health mandate. It represents a socioeconomic and urban resilience strategy. Safe street food supports employment stability, tourism attractiveness, and consumer confidence in informal food systems. The data generated in this study provide an empirical foundation for developing predictive risk models that capture the seasonal and geographic variability of Lebanese street food sector. A coordinated, science-driven policy framework, aligned with the Codex Alimentarius and the WHO One Health approach, could transform Lebanese vibrant street food economy into a model of safe, sustainable, and inclusive urban food systems.

A critical insight emerging from this work is the absence of internationally recognized microbiological standards tailored explicitly to street-vended foods. Neither Codex Alimentarius, EFSA, nor WHO currently defines numerical microbial limits for foods prepared and sold in informal settings, leaving countries to rely on general ready-to-eat food criteria that do not reflect the realities of street-food environments. This regulatory vacuum places greater responsibility on national authorities to establish contextualized, evidence-based guidelines that address the unique operational, environmental, and behavioral factors shaping food safety in these settings. By providing robust, seasonally grounded microbiological data, this study offers a scientific foundation for Lebanon to develop national street-food standards, strengthen municipal food-safety systems, and advance a One Health approach to protecting consumers in rapidly evolving urban food markets. Ultimately, safeguarding the microbiological quality of street-vended foods in Lebanon and across the broader Mediterranean region will require coupling scientific evidence with proactive governance, ensuring that this vibrant and economically vital sector evolves into a model of safe, resilient, and inclusive urban food systems.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

CFI, RAB: Formal analysis, Software, Methodology; CFI, MGA: Validation, Methodology; CFI, RAB, NG, LM: Investigation, CFI, RAB, NG, LM, MGA: Writing–review & editing, CFI, MGA: Supervision; CFI Funding aquisition.

Funding

This work was supported by the American University of Beirut Research Board (Award# 104514).

Data availability

All data generated or analyzed during this study are included in this published article, and its supplementary information files.

Declarations

Competing interests

The authors declare no competing interests.