Food safety knowledge and practices of stationary street food vendors: a case study of grassy park, city of Cape Town

1 Introduction

In the Global South, street-vended foods constitute a vital socio-economic lifeline, providing affordable nutrition to rapidly expanding urban populations while sustaining the livelihoods of millions of informal entrepreneurs. Beyond merely providing sustenance, this informal economy serves as a crucial buffer against poverty and unemployment, particularly for women and marginalized urban groups (Mahopo et al., 2022; Rakha et al., 2022). However, in South Africa, this essential sector operates at a precarious intersection of food security and public health (Nkosi and Tabit, 2021).

The World Health Organization (WHO) estimates that foodborne diseases result in millions of illnesses annually, with the African region bearing a disproportionate share of this burden (World Health Organization, 2024; Grace, 2023). In street food environments, pathogens such as Escherichia coli, Salmonella, and Staphylococcus aureus are frequently transmitted due to inadequate temperature control and poor handling practices (Viana et al., 2025; Nkosi and Tabit, 2021). The catastrophic 2017–2018 listeriosis outbreak in South Africa, the largest recorded globally, resulting in 216 fatalities, starkly underscored these systemic vulnerabilities within the national food supply chain, highlighting the urgent need for rigorous safety protocols across both the formal and informal sectors (Kaptchouang Tchatchouang et al., 2020).

Despite the implementation of South Africa’s Regulation R638, which mandates strict hygiene standards for all food premises, compliance among street traders remains notoriously low (This non-compliance is frequently attributed to a “knowledge-practice gap,” where a theoretical understanding of hygiene fails to translate into safe food handling behaviors (Desye et al., 2023). Current literature utilizing the knowledge-attitude-practice (KAP) model reveals that while many South African vendors possess a baseline awareness of personal hygiene, such as basic handwashing, significant cognitive and behavioral deficiencies remain (Majova, 2024). Studies indicate that up to 76% of vendors exhibit poor overall food safety knowledge, particularly concerning critical safety limits, internal cooking temperatures, cross-contamination, and the safe storage of ready-to-eat foods. Furthermore, these behavioral gaps are severely exacerbated by structural and environmental barriers (Oladipo-Adekeye, 2020; Nkosi and Tabit, 2021).

Nationwide, up to 70.8% of vendors in certain regions are forced to operate without access to potable running water or adequate waste disposal facilities (Haves, 2021). In Cape Town specifically, research highlights that a vast majority of informal food stalls lack basic sanitation necessities, such as food-safe sanitizers, washbasins, or clean drying cloths (Rohith, 2021; Hill et al., 2018). Grassy Park, a busy suburban node in Cape Town characterized by high daily commuter footfall and a thriving informal trading sector, exemplifies these compounding challenges. For stationary vendors in this area, the public health risks are further amplified by an over-reliance on aging infrastructure and inconsistent municipal services. By evaluating the food safety landscape in Grassy Park, this study addresses a critical regional data void. Ultimately, the findings will provide the empirical evidence necessary for municipal health authorities to design targeted, localized interventions that safeguard consumer health without compromising the economic survival of informal traders.

2 Methodology

2.1 Research study and settings

This study employed a quantitative, cross-sectional descriptive design. Research was conducted in Grassy Park, a suburb of Cape Town in South Africa’s Western Cape Province. Geographically, Grassy Park is bounded by Pelican Park to the east, Ottery and Plumstead to the north, and Retreat to the west, with the False Bay Coastal Park situated to the south. Because it serves as a major transit hub for commuters traveling across the City of Cape Town, Grassy Park is a high-traffic area particularly conducive to street food vending.

2.2 Study population and sampling

The study population comprised male and female street food vendors aged 21 years and older, regardless of nationality or employment status. The inclusion criteria focused specifically on vendors selling ready-to-eat foods, many of which are considered high-risk due to their composition and handling requirements. These included boerewors (sausage) rolls, pies, chicken fillet burgers, beef burgers, whole-grilled chicken, and vetkoek with mince curry. Food handlers who served as assistants or helpers were excluded from the study; therefore, the assessment of food safety knowledge and practices was limited to primary vendors, and no data were collected from support staff during either the pilot or the main study phases. According to the City of Cape Town’s Environmental Health Department, approximately 200 vendors operate within the Grassy Park area. Due to the transient nature of these vendors and the lack of a formal registry of their exact locations, a convenience sampling method was employed to ensure practical access to the study participants.

The study’s sample size was determined using the Yamane formula (Yamane, 1967), in which “N” denotes the estimated population size (200) and “e” signifies the margin of error (0.005). The calculated expected sample size was 133. However, the actual sample comprised 108 vendors, representing approximately 54% of the estimated local vendor population. Although this figure is slightly below the ideal statistical calculation of 133, the sample size was deemed sufficient to yield a statistically significant multivariate model (p < 0.001), which accounted for 26.1% of the variance in food safety practices.

n = N 1 + N e 2

2.3 Socio-demographic variables

Data was collected using a structured questionnaire specifically designed and adapted from the Food and Agriculture Organization (FAO) and World Health Organization (WHO) guidelines for assessing food safety, as well as South Africa’s Regulation R638 requirements. The instrument consisted of four primary sections: Section A gathered socio-demographic and operational data (e.g., age, gender, education level, and alternative income sources), while the other sections included Section B: food safety knowledge, Section C: compliance with food safety laws, and Section D: food safety practices.

Food Safety Knowledge: This section employed twelve closed-ended questions to evaluate participants’ understanding of food poisoning, contamination, temperature control, handwashing with soap, and sanitation practices. Responses were scored using a binary system (1 for “Yes,” 0 for “No”), with individual totals ranging from 0 to 12, where higher scores indicated a greater level of theoretical proficiency.

Regulatory Compliance: This section utilized a 14-item observational checklist to assess adherence to South African hygiene standards, including the possession of a Certificate of Acceptability and the maintenance of clean equipment. Compliance was scored as 1, while non-compliance was scored as 0, resulting in an aggregate score ranging from 0 to 14 for each participant.

Food Safety Practices: This 10-item observational section monitored real-time hygiene and food handling practices, encompassing hot-holding temperatures (maintained above 65 °C), pest management, and waste disposal. Practices were scored as one for “yes” and 0 for “no,” yielding total scores ranging from 0 to 10, thereby quantifying the overall practice score. Individuals who scored above 6 were classified as having demonstrated good practices, while those who scored 5 or below were deemed to have exhibited poor practices.

2.4 Pilot study

A pilot study was conducted with 10 street food vendors in Athlone CBD, Cape Town, a setting comparable to Grassy Park in terms of high commuter activity and informal trading conditions. Conducted between 15 and 30 November 2023, the pilot followed the same ethical procedures as the main study. The pilot assessed the clarity, relevance, and internal consistency of the questionnaire. Prior to piloting, content validity was established through expert review by the research supervisor, an Environmental Health specialist, ensuring alignment with FAO, WHO, and national regulatory guidelines. Minor refinements were subsequently made to improve question clarity and coherence. Participants required minimal clarification, indicating acceptable face validity. The instrument was feasible to administer, requiring approximately 15–20 min per participant. Data from the pilot were not included in the final analysis.

2.5 Data analysis

The study’s analysis plan was structured into two distinct phases, beginning with a descriptive analysis designed to characterize the 108 participants and their environment using frequencies (n) and percentages (%) for socio-demographic profiling, alongside measures of central tendency and dispersion, specifically means, medians, and standard deviations (SD), to identify specific gaps in food safety knowledge and compliance. This was followed by an inferential analysis phase aimed at determining the underlying drivers of food safety practices through a three-step progression: first, using Pearson correlation coefficients (r) to measure linear relationships between variables; second, employing bivariate regression to identify the unadjusted effects of individual factors on practices; and finally, utilizing a Multiple Linear Regression model to isolate significant predictors while controlling for all other variables. By calculating coefficients (β), p-values, and 95% Confidence Intervals, this multivariate approach enabled the identification of the true impact of factors such as higher education and age, while highlighting critical anomalies, such as the unexpected negative impact of current training, for further investigation.

3 Findings

3.1 Participants’ socio-demographics

The socio-demographic profile of the participants is summarized in Table 1. The sample was predominantly female (n = 69; 64%), with the largest age cohort being those aged 31–40 years (n = 45; 42%), followed by the 21–30 range (n = 42; 39%). Regarding socio-economic indicators, 56% (n = 60) of respondents relied solely on street food vending as their primary source of income. Educational attainment was generally low, with the vast majority (n = 105; 97%) holding a Grade 12 qualification or less. Notably, only one-third of the participants (n = 36; 33%) reported having attended formal food safety or hygiene training.

TABLE 1. Socio-demographics of participants.

3.2 Knowledge of food safety among participants

Among the 108 participants (N = 108), the mean knowledge score was 11.03 (SD = 2.08), with a median of 11.00, indicating a relatively symmetrical distribution of scores. Figure 1 presents the distribution of food safety knowledge among the study participants. More than half of the participants (n = 66; 61%) understood what food safety means, achieving a mean knowledge score of 12.18 (SD = 1.38). Additionally, a significant number of food handlers in the study (n = 96; 89%) recognized what food poisoning is, with a mean knowledge score of 11.41 (SD = 1.86). The results also show that 78 participants (72%) understood food contamination, with a mean knowledge score of 11.81 (SD = 1.70). Most participants (n = 105; 97%), with a mean knowledge score of 11.09 (SD = 2.11), believe that handwashing, cleaning, and sanitizing are important during food preparation and handling. All participants (n = 108; 100%) demonstrated awareness that food must be kept covered during display and storage, and that they should not work with food if they are sick or injured, achieving a mean knowledge score of 11.03 (SD = 2.08). The majority of participants (n = 75; 69%) understood the concept of temperature control for food, with a mean knowledge score of 11.84 (SD = 1.73). However, less than half of the participants knew the correct storage temperatures for hot and cold foods. Specifically, 42 participants (39%) indicated that cold foods must be kept below 5 °C, with a mean knowledge score of 10.71 (SD = 1.69), while 39 participants (36%) stated that hot foods must be kept above 65 °C, achieving a mean knowledge score of 10.46 (SD = 2.23).

3.3 Food safety compliance with relevant legislation (City’s by-laws)

Figure 2 shows the distribution of Food Safety Compliance among the study participants concerning the City’s by-laws. Only a few vendors (n = 27; 25%) possessed a valid certificate of acceptability, with a compliance mean score of 8.56 (SD = 4.353). Most vendors had visually clean non-product contact surfaces and equipment/utensils, with n = 105 (97%) showing no dirt accumulation (mean score 7.57, SD = 3.305) and n = 102 (94%) having clean equipment/utensils (mean score 7.65, SD = 3.323). Less than half of the vendors (n = 51; 47%) used adequate cleaning equipment (mean score 10.35, SD = 2.568), while approximately half (n = 57; 53%) had clean and well-maintained cleaning equipment (mean score 9.53, SD = 3.328). Fewer than half (n = 48; 44%) stored cleaning equipment in a clean area (mean score 10.56, SD = 2.500). Additionally, n = 54 (50%) of vendors correctly used and stored wiping cloths (mean score 9.78, SD = 3.166). Most vendors exhibited good housekeeping and clean-as-you-go practices, with n = 96 (89%) showing good housekeeping (mean score 7.91, SD = 3.261) and n = 81 (75%) performing clean-as-you-go practices (mean score 8.33, SD = 3.343). Interestingly, n = 48 (44%) had access to potable running water (mean score 10.31, SD = 2.714). None of the vendors (n = 0; 0%) had cleaning instructions available. Very few vendors (n = 15; 14%) had cleaning chemicals available and in use (mean score 13.20, SD = 0.414), with only n = 6 (6%) having adequately marked cleaning chemicals (mean score 13.50, SD = 0.548).

3.4 Correlation between socio-demographic, knowledge, and compliance

The Pearson correlation analysis (Figure 3) revealed several significant socio-demographic and performance-based relationships. Notably, food safety or hygiene training showed a significant positive correlation with Knowledge Scores (r = 0.388, p < 0.01), highlighting the effectiveness of formal instruction. In contrast, gender was significantly negatively associated with knowledge (r = −0.400, p < 0.01) but positively associated with Compliance Scores (r = 0.238, p < 0.05). Economic stability also played a role, as having alternative income sources was negatively correlated with compliance (r = −0.318, p < 0.01), suggesting that full-time vendors may prioritize regulatory adherence more than those with diversified income. Finally, although a positive trend was observed, the correlation between knowledge and compliance did not reach statistical significance (r = 0.126, p > 0.05), indicating that increased knowledge alone may be insufficient to ensure improved food safety practices in the vending environment.

3.5 Participants’ practices and factors influencing them

Statistically significant differences in food safety practices were observed between vendor groups as shown in Figure 4. While basic environmental compliance was achieved through the universal availability of waste bins (100%) and the consistent covering of products (100%), severe deficiencies were identified in high-risk behavioral areas. Specifically, 67% of vendors failed to maintain hot-holding temperatures above the required 65 °C (p = 0.003), a failure that directly mirrors the 36% microbiological contamination risk observed across the sample (p = 0.007). These behavioral gaps are further exacerbated by significant structural barriers; despite 64% of staff attempting effective hand-washing practices (p = 0.905), 94% of vendors operated without a dedicated hand-washing facility near the food handling area (p = 0.088). Furthermore, the study identified a statistically significant disparity in environmental hygiene, where 97% of sites showed evidence of pest activity (p = 0.010) and 14% failed to maintain visually clean storage containers (p = 0.005). Access to basic sanitation was also a critical determinant of performance, as nearly half of the vendors (47%) operated without nearby toilet facilities (p = 0.003). Collectively, these findings underscore a profound “Knowledge-Practice Gap,” where even in the presence of basic waste infrastructure, critical safety limits and sanitation requirements remain largely unaddressed due to a lack of essential facilities and behaviorally-oriented training.

The bivariate analysis in Table 2, the findings indicate that the majority of socio-demographic factors do not demonstrate a statistically significant individual relationship with the examined practices, with several notable exceptions. The most significant observation pertains to food safety hygiene training, where individuals who received such training scored 3.139 units lower than those who did not. Furthermore, income was associated with a slight positive increase in scores among individuals with an income relative to those without. In contrast, factors including gender, age, educational qualifications, as well as both knowledge and compliance scores, did not achieve statistical significance, as their p-values exceeded the 0.05 threshold and their 95% confidence intervals encompassed zero. Nonetheless, the knowledge score indicated a marginal positive increase in practice scores. This implies that, in the absence of controlling for the interrelated effects of other variables, these specific socio-demographic characteristics do not appear to serve as robust independent predictors of the outcome within the confines of this model.

TABLE 2. Multivariate regression analysis of factors influencing the practices.

Coeff- Coefficient, Std err-standard error.

represents statistically significant p-value at 0.050 or less.

The final multivariate regression model analysis (Table 2), conducted with a sample of 108 observations, indicates that the model is statistically significant (F (9, 98) = 3.85, p < 0.001) and explains approximately 26.1% of the variance in practice scores (R² = 0.2613). Several factors were identified as significant predictors of the outcome. Notably, individuals with a Post-Matric Qualification exhibited a substantial increase in their scores (β = 16.061, p = 0.004), and those within the 31–40 age group demonstrated significantly superior performance compared to the reference group (β = 7.402, p < 0.001). Additionally, higher Knowledge scores were positively associated with improved practices (β = 1.049, p = 0.027). Conversely, being Male (β = −4.151, p = 0.044) and, interestingly, having received Training (β = −10.403, p < 0.001) were associated with lower practice scores. Other variables, including older age categories, income status, and compliance scores, did not demonstrate a statistically significant effect on the dependent variable.

4 Discussion

The findings of this study reveal a complex interplay between socio-demographic factors and food safety outcomes among stationary street food vendors in Grassy Park. A significant positive correlation was observed between food safety training and knowledge scores, reinforcing the established consensus in global literature that formal instruction is a primary driver of theoretical proficiency (Adesokan et al., 2015; Rifat et al., 2022). However, the study uncovered a notable “gender paradox” where gender was negatively associated with knowledge but positively correlated with compliance scores.

This suggests that while certain demographics may score lower on formal assessments, they may possess superior practical habits, potentially rooted in domestic food handling experience, a phenomenon noted in several developmental contexts (Gebre et al., 2023). Furthermore, the negative correlation between alternative income sources and compliance suggests that “professionalism” and livelihood dependency are critical drivers of safety; vendors whose sole income is derived from their stall may be more risk-averse regarding regulatory adherence.

Crucially, the lack of a statistically significant relationship between knowledge and compliance confirms the “Knowledge-Practice Gap” frequently cited in food safety research (Sheehama and Singh, 2025; Magwe, 2024; Boakye et al., 2023). This indicates that increasing a vendor’s theoretical understanding is, in isolation, insufficient to change behavior, likely due to external structural barriers such as limited access to potable water or sanitation infrastructure in the vending environment (Boakye et al., 2023; Clayton et al., 2002).

The finding that Post-Matric Qualifications and Knowledge are significant drivers of food safety practices, while traditional Training shows a paradoxical negative correlation, suggests a critical disconnect between formal instruction and actual behavior among street vendors in Grassy Park. This “knowledge-practice gap” is frequently explored in the literature, highlighting that while cognitive understanding is a prerequisite, it does not always translate into habits in high-pressure, resource-limited environments (Alyousef et al., 2025; Wallace et al., 2022).

Specifically, the negative coefficient for training (β = −10.403) may reflect “ritualized learning,” where vendors attend mandatory sessions to meet regulatory requirements without internalizing the behaviors, or it may reflect that the training provided lacks the practical, site-specific nuance required for the informal sector. These findings align with socio-demographic trends observed in other urban African contexts, where female vendors and those in the “prime” working age (31–40) exhibit greater caution, likely due to a combination of domestic experience and a greater perceived risk of losing their livelihood (Magqupu et al., 2024; Akabanda et al., 2017).

Another specific study that supports these nuanced findings is da Vitória et al. (2021), titled “Food safety knowledge, attitudes and practices of food handlers: A cross-sectional study in school kitchens in Espírito Santo, Brazil.” While their study focused on a different geographic region, their findings revealed similar discrepancies that mirror those in Grassy Park. They found that while food handlers often possessed high theoretical knowledge scores, their actual observed practices were significantly lower.

Furthermore, da Vitória et al. (2021) identified that experience, education, and training time were consistent predictors of better hygiene practices, supporting this study’s finding that a Post-Matric Qualification is a powerful determinant. Based on these results, appropriate solutions are needed to guide future research and improve training and guidance on food safety. Food handlers must be provided with information to implement in their work routines.

One of the study’s strengths is the provision of critical empirical evidence regarding the “knowledge-practice gap” within the informal food sector of Grassy Park. Through the integration of structured questionnaires and direct observational checklists, the research uniquely identifies significant socio-demographic predictors, including age and educational attainment, while simultaneously highlighting urgent structural deficiencies in sanitation infrastructure. These insights offer a robust foundation for municipal authorities to design targeted, behaviorally oriented public health interventions. However, the study’s limitations include the use of a convenience sampling approach and focus on a single geographical area, which may restrict the generalizability of the findings. Additionally, reliance on self-reported data may introduce response bias, and the exclusion of support staff, along with the cross-sectional design, prevents the establishment of long-term causal relationships.

The strengths of this study stem from its methodological triangulation, utilizing structured questionnaires and direct observation to provide empirical evidence of the “knowledge-practice gap” among street vendors. By identifying significant predictors such as age and education through multivariate analysis, the study offers actionable insights for municipal health interventions.

5 Conclusion

The examination of street food vendors in Grassy Park reveals a significant disconnection between theoretical knowledge and practical application. While formal education enhances a vendor’s understanding of food safety protocols, the persistent “Knowledge-Practice Gap” shows that awareness of regulations does not automatically lead to compliance. The findings indicate that food safety behaviors in this context are primarily influenced by socio-economic factors and experiential learning rather than traditional classroom instruction. Key insights include:

• The Training Paradox: Conventional training programs may be overly generic or “ritualized,” failing to provide vendors with the site-specific, practical tools needed to change behaviors in high-pressure situations.
• The Gender and Experience Factor: Compliance with food safety practices is significantly influenced by domestic experience and demographic factors, with female vendors often demonstrating better adherence despite lower scores on formal knowledge assessments.
• Professionalism as a Driver: The highest levels of compliance are observed among individuals whose primary income comes from their stall, suggesting that the perceived risk of business loss is a stronger motivator than regulatory compliance alone.
• Structural Barriers: The gap between knowledge and practice highlights external constraints (such as limited access to potable water and sanitation infrastructure) that prevent even the most informed vendors from achieving optimal safety standards.

Ultimately, these findings suggest that improving public health outcomes in the street food vending sector requires moving beyond purely cognitive, theory-based training approaches. Future interventions should focus on behaviorally oriented coaching and infrastructure development, addressing the contextual and economic challenges that currently hinder safe food handling practices.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Fernando Perez Rodriguez, University of Cordoba, Spain

Reviewed by: Alekhya Sabbithi, Health Researcher, United States

Rashmi Savant, Indian Council of Medical Research (ICMR), India