Evaluating the impacts of school garden-based programmes on diet and nutrition-related knowledge, attitudes and practices among the school children: a systematic review

The search was conducted between 11th November 2020 to 6th February 2021. Five databases were used, including PubMed, Embase, Cochrane, Web of Science and Scopus for primary research articles published from year 2000. This timeframe was chosen with the aim of obtaining the most recent SGBP intervention studies. The following search terms were used: (1) school children as the targeted population: adolescent* OR boy? OR child* OR children OR girl? OR juvenil* OR kid? OR preschool* OR school* OR teen* OR youth* OR young OR “school children” OR student*; (2) school setting: school* OR nurser* OR kindergarten* OR kindergarden*; (3) garden-based interventions: garden* OR gardening OR plant* OR fruit* OR vegetable* OR “fruit vegetable*” OR “fruit growing” OR “vegetable growing” OR seed* OR tree* OR “organic agriculture” OR “organic farming” OR “organic food” OR farm; (4) outcome measures on diet and nutritional related KAP: (eating OR diet* OR food OR dietary OR nutrition OR nutritional OR fruit* OR vegetable*) AND (knowledge OR attitude OR practi?e* OR behavio?r* OR preference* OR habit* OR intake* OR consumption* OR healthy OR skill* OR pattern* OR diversity OR diverse OR perception*) OR “energy intake” OR “appetite” OR “portion size*” OR “food fussiness” OR “food neophobia”; (5) study design: “controlled trial*” OR “intervention” OR randomised OR randomized OR trial* OR “randomised controlled trial*” OR “randomized controlled trial*” OR follow-up stud* OR program evaluation*” OR “controlled before-after stud*”. Details of the search strategies used for each database are presented in the Supplementary Table 1.

Studies were included if they met the following inclusion and exclusion criteria.

Garden-based interventions that did not organise by the school such as community-based gardening programmes, community youth interventions, summer holiday extra-curricular activities or clubs were excluded. Study organised by the school but occurred at the community level such community gardens, however, were included as the participants were still being regarded as “students”. Interventions with only teaching gardening related knowledge without actual hand-on gardening component were excluded. Studies that did not regard school gardening as their primary intervention or did not specify the age of participants were also excluded. Studies that only focused on describing school-based gardening programme without addressing its effects on nutritional KAP were excluded. Editorials, commentaries, opinions, review articles and observational studies such as cross-sectional studies, prospective and retrospective cohort studies were not included as well as unpublished, grey literature and ongoing studies with only preliminary findings.

Studies obtained from the search were uploaded to ENDNOTE (X7, Thomson Reuters). Screening and selection of studies for inclusion in this review were performed by a reviewer and the decisions were checked by the other reviewer. During the first round of screening, the title and abstract were checked for eligibility based on the inclusion and exclusion criteria. In the second round of screening, full-text articles were obtained and screened for eligibility using the same criteria. Disagreement between reviewers was resolved by discussion and by a third reviewer.

A standardized data extraction form was utilized to obtain the following information, where possible: author, year of publication, journal source, source of funding, study design, year of study, country or population, sample characteristics (e.g., gender, age, socioeconomic status etc.), sample size, intervention size, control size, intervention group description (activities included in the SGBP), control group description, duration of intervention, outcome measures (e.g., indicators related to KAP around diet and nutrition) and main findings. Any disagreements were resolved through discussion with the research team.

A systematic review synthesising the qualitative evidence of school garden-based programmes was conducted. The findings on the impact of school garden-based programmes in affecting school children’s KAP around diet and nutrition were reported according to the components of the interventions via categorising them into school garden-based programmes with and without parental involvement. A meta-analysis on any of the quantitative data extracted was unable to be performed due to the heterogeneity and variation in the study design, outcome measurement and intervention component.

The risk of bias of the individual studies included was independently assessed by two reviewers. Any disagreement on the risk of bias between reviewers was resolved by discussion and by a third reviewer when necessary. The risk of bias of the individual studies included was assessed using the Academy of Nutrition and Dietetics, Quality Criteria Checklist [29]. The 10 questions focus on (1) how clear the research question was; (2) selection of participants; (3) randomization/ group comparability; (4) description of withdrawals; (5) how the blinding was; (6) whether study procedures were described clearly; (7) whether the outcomes were clearly defined; (8) were appropriate statistical analyses applied; (9) did the results support the conclusion; (10) funding or sponsorship bias. To be rated positive, each of the criteria 2, 3, 6 and 7 must be met and the majority of 10 criteria overall. Any of criteria 2, 3, 6 and 7 not being met resulted in a neutral rating. If most criteria (i.e., more than 6 of them) were not met, the article would have a negative rating.

In total, 25,726 school children recruited from 341 schools and 8 nursery centres from 12 different countries were included in this review. Most of the studies were mainly reported from the populations of the United States (n = 18). Six studies were conducted in low- and middle-income countries (LMICs) (Nepal, Bhutan, Burkina Faso and Brazil) and 29 studies were conducted in developed countries (United States, Australia, United Kingdom, Portugal, Canada, South Korea, Netherlands and Belgium). Sample size ranged from 1 to 49 schools and 44 to 4300 participants, with more than 80% of the included studies recruited more than 100 participants (n = 29/35). Participants’ age ranged from 2 to 19 years old, with the majority from the age of 8–12 years old. Duration of intervention ranged from 6 weeks to 4 years (mean ± SD: 10 ± 11 months) and integrated school gardening intervention activities included outdoor or indoor classroom gardening (e.g., Earthbox gardening); harvesting lessons; cooking lessons and experimental kitchen activities utilising harvests; taste tests; nutrition-related education on food cultivation, healthy living skills, agriculture and nutrition science; physical education; healthy F&V snack program; poster, poem and nutrition and vegetable charts displays on school boards, meat-free Monday, using locally source produce in school meals and market days to sell produce from the garden and local farmers’ market visit. Outcomes of each study varied, but the majority primarily focused on the changes in children's KAP on food consumption (particularly F&V).

The quality appraisal of the studies included is reported in Fig. 2. Almost half of the studies included had a low risk of bias with the remaining rated unclear risk (neutral). No study included had a high risk of bias. Categories that were commonly rated as weak (e.g., with more than half of the studies rated a high risk of bias) were statistical analysis, blinding and withdrawal description. Most of the studies (n = 33/35) failed to apply appropriate statistical analysis, studies rated as low risk of bias in this category were able to address the confounding factors as well as the application of intention to treat analysis. Majority of the studies (n = 32/35) failed to describe the allocation concealment or blinding of researchers, participants, or data collectors. In addition, a large proportion of studies did not describe the method of handling withdrawals (n = 24/35), including the follow-up method and withdrawal reasons. Detailed quality appraisal of each study is reported in Supplementary Table 2.

A total of 35 studies have been included, and the characteristics of each intervention study are reported in Table 1. There were 18 out of the 35 studies in which the intervention included parental involvement. In this review, level of parental involvement differed between studies, ranging from students gardening with parents; student and family cooking events; parent gardening, home gardening, maintenance of school garden, school visit invitation to receive a brief of school gardening project, end-of-programme celebration invitation, take-home materials (e.g., “Family Stories” booklet and recipe cards) and parent newsletter (considered as weak parental component or low activity intensity). The main findings of the impacts of SGBP on intervention outcomes and the study quality are reported in Table 2.

The impacts of school garden-based programmes with or without parental involvement on the children’s diet and nutritional-related knowledge, attitudes, and practices from the 35 studies included are summarised in Fig. 3. Non-significant increase is regarded as no change in terms of the effectiveness on improving the measure outcomes as reported by the studies.

Children’s F&V intake was the most studied outcome (n = 26). Six out of 10 studies demonstrated SGBP without parental involvement, with a shorter intervention duration ranging from 12 weeks to 28 weeks and a smaller sample size ranging from 77 to 320 participants, resulting in a more favourable outcome on children’s vegetable intake, especially among the younger children from pre-schools and primary schools [22, 32, 37, 47, 57, 59]. Contrarily, most of the SGBP with parental involvement did not show significant improvement in children’s vegetable intake (n = 11/16) [26, 31, 35, 38, 40, 44, 45, 52, 54‐56]. However, this may be due to the longer intervention duration ranging from 1 year to 4 years, larger sample size ranging from 89 to 4300 participants or intervening at an older age (e.g., secondary school-aged children). Similar findings were observed in children’s fruit intake, SGBP with a shorter duration (~ 12 weeks) and smaller sample size (~ 77 to 99 participants) showed better improvement in children’s fruit intake among the preschool and primary school-aged children [47, 59]. However, the majority of the SGBP with longer intervention duration (~ 1 year to 4 years) and larger sample size (~ 60 to 4300 participants), did not observe significant improvement in children’s fruit intake, regardless of the parental involvement [3, 26, 27, 30, 31, 36, 38, 40, 42, 44, 45, 52, 54‐56].

Four studies reported the positive impacts of SGBP on dietary fibre, and a study showed increased intake on vitamin A and vitamin C at the post-intervention [37, 44, 55, 59]. A study conducted on the Portuguese population showed a promising effect of SGBP in reducing students’ salt intake [51], and the other three studies found no significant improvement in reducing sugar-sweetened beverages (SSB) and ultra-processed food consumption at the post-intervention [30, 38, 49]. In addition, a small number of studies investigated the impact on the variety of fruits (n = 6) and vegetables (n = 7) consumed, with the majority not able to demonstrate a significant improvement (n = 5/6 [26, 27, 40, 42, 48]; n = 4/7 [26, 27, 40, 48], respectively).

Nutrition, gardening, agricultural and science-related knowledge was the second most studied outcome (n = 20). Fourteen out of 20 studies reported that SGBP with or without parental involvement demonstrated significant improvement in children’s nutritional knowledge at the post-intervention, especially those shorter SGBP interventions (less than a year) integrating with classroom education and intervening at a younger age (6 to 15 years old) [3, 22, 26, 32, 35, 39‐41, 46, 49, 50, 56‐58]. It is worth highlighting that high sample size variability has been observed in those studies that have reported changes in children’s nutritional knowledge.

Two thirds of the reported studies showed significant improvement in children’s attitudes and behaviours towards vegetables at post-intervention (n = 13/19) [22, 26, 27, 32, 33, 35, 41, 47, 49, 52, 56‐58]. Parental involvement in SGBP seems to produce better improvement in children’s attitudes towards vegetables, especially those with shorter intervention duration ranging from 12 weeks to 1 year, regardless of the sample size and the children’s age group (n = 7/9) [26, 27, 33, 35, 49, 52, 56]. Similar findings were observed on children’s attitudes towards fruits, the majority of the SGBP with parental involvement reported improvement on children’s attitudes towards fruits (n = 3/5) [26, 27, 52] compared to those without parental involvement (n = 3/7) [21, 47, 57]. In addition, parental involvement in SBGP seems to exert beneficial effects on improving children’s willingness to consume F&V, especially when intervened at a younger age (aged 3–12 years old) with an intervention duration ranging from 8 weeks to 2.5 years (n = 4/6) [48, 52, 54, 56].

Four studies reported on the dietary self-efficacy in children, which refers to children’s self-belief in their food knowledge level and the self-confidence to purchase, plan, prepare and cook food, as well as to make appropriate food decisions to achieve higher nutritional value [60]. Most of the studies reported no significant improvement in dietary self-efficacy (n = 3/4) [39, 46, 58].

Eight studies reported the impact of SGBP on home F&V availability and consumption [21, 22, 34, 35, 43, 45, 48, 49]. Only one study from the US reported positive findings on home vegetable availability, with the remaining failing to demonstrate significant improvement [43]. In addition, one study from the US investigated the effect on school F&V availability reported a positive finding on improving school F&V availability [3].

Despite the proposal of the knowledge, attitudes and behaviour communication model by Contento et al., 1992 which suggested a linear positive association between the three components that potentially influence the “practice” of such behaviour, recent research argued that such a relationship is far more reciprocal and dynamic [61, 62]. It is thus essential to understand the impacts of SGBP in addressing such association to maximise its intervention effect towards improving children F&V intake to alleviate childhood malnutrition. In this review, the findings generated from the 35 studies included indicated that school garden-based programmes were effective in increasing diet and nutritional knowledge, as well as promoting positive attitudes and behaviours (acceptability) towards vegetables among the school children, however, most of the studies reported no significant in their dietary practices such as F&V consumption and dietary diversity.

In this review, SGBP have shown promising effects in improving children’s knowledge of food, nutrition, gardening and science (Fig. 2). Acquisition of knowledge is the basis for behavioural change [63]. Active participation in school gardening activities in combination with in-class food and nutrition curriculum has strengthened not only children’s horticulture skills but also children’s declarative knowledge (what is a healthy diet), procedural knowledge (how to achieve a healthy diet) and conditional knowledge (when and why healthy diet) [41, 61, 64]. Children thus have a higher potential and ability to make better and healthier food choices. In addition, children have demonstrated a greater ability to identify unfamiliar food, as less typical vegetables are intentionally emphasized in class during the intervention and incorporated into cooking activities and recipes [37]. The “seed to mouth” nature of the cooking and gardening programme, where children eat what they have grown in the school garden, also explains why children are more likely to recognize the types of crop they have consumed [37].

SGBP may encourage children’s attitudes and acceptability to consume new food and reduce food neophobia via increasing their F&V exposure. Food neophobia refers to the reluctance to eat and try novel food, which often acts as a barrier to promote F&V preference and consumption among children and contribute to the development of unhealthful food habits [65‐67]. In this review, SGBP were successful in improving children’s willingness to consume F&V. The direct involvements in growing and cooking own garden produces were associated with an increase in children self-reported willingness to consume new food [52, 56]. Additionally, the nature of the kitchen gardening programme, where children are encouraged to freely taste and share self-prepared meals with each other during kitchen class with no pressure to eat, has created a favourable social environment for children to try unfamiliar food and potentially reduce their food neophobic rate. A study conducted by Morgan et al., 2010 highlighted an increase in willingness and preference towards vegetables not only in the vegetables grown in the school garden, but also those in general, suggesting the intervention was successful in exerting influence beyond scope of the school garden, and even extended to those children did not directly expose to [56].

Interestingly, improvements were observed in children’s attitudes and acceptability towards vegetables, but not for the fruits. Previous evidence suggested that being actively involved in the food production and preparation process may exert positive influences on foods that are particularly hard to change preference towards [68, 69]. The hands-on learning experience provided by SGBP offers children regular positive exposure to vegetables. Through direct experience, for instance, crops growing, harvesting as well as food preparation, children increase familiarity with vegetables, and thus more likely to positively accept and improve taste preference towards such food [70‐72]. This might also explain the insignificant intervention effect on improving children’s attitude towards fruits as the crops grown in the interventions were dominated by vegetable species, children thus had comparatively fewer opportunities to interact with fruits and thus lowering the chance to increase their acceptability towards such food.

Intervention effects on improving F&V dietary intake remain inconclusive due to the mixed results generated from the studies. F&V knowledge is one of the most important determinants of their consumption [46, 73‐76]. This is also supported by the social cognitive theory as augmentation in food- and nutrition-related knowledge with the acquisition of horticultural skills could increase the behavioural capacity regarding F&V intake [60, 77]. Knowledge, attitudes, taste preferences and acceptability are often being described as one of the strongest predictors for future F&V intake [21, 37, 46]. However, in this review, our finding demonstrated that SGBPs improvement in such predictors may not be sufficient to translate into an actual increase in F&V uptake.

The weak intervention effect on dietary practices found in this review denotes SGBP might fail to address other determinants that might exert a greater influence on children’s F&V consumption. This includes school and home F&V availability, parental food habits and feeding practices, the level of perceived behaviour control (e.g., ease of increasing F&V intake) and dietary self-efficacy level (i.e., self-confidence in being able to increase F&V intake) among individuals, as well as the peers’ influences [78]. Children spend most of the time learning and working as a team at school, they are more likely to follow the group perspective and consider less about their own attitudes and beliefs [79]. In other words, when the peers are more health-conscious, individuals are more likely to consume more healthy food such as F&V regardless of their own attitude towards such foods [79, 80].

Nonetheless, studies with positive findings on improving dietary practice found that SGBP was helpful in building a “personal connection” between children and the crops they have grown through direct experience on crop planting and nurturing. The “garden-grown” nature also adds an extra value to the F&V, children are thus more inclined to try and consume those healthful foods [37, 58]. An increase in school F&V availability during intervention also acts as a drive-in promoting children’s F&V consumption [22]. Moreover, SGBP can exert its influence beyond the school setting, as it is found higher varieties of vegetables not cultivated from the school garden were consumed [22]. However, the studies with positive findings may be prone to bias. For instance, the study conducted by Parmer et al., 2009 may be prone to gender bias as it was conducted predominantly in males [57]. Previous studies reported gender differences in F&V intake with the female being more likely to consume F&V and have greater perceived behavioral control and favorable attitudes towards F&V [81‐84]. The positive result from McAleese et al., 2007 might be prone to measurement error, as a one-time 24-hour food recall was used [59]. The use of one-time 24-hour recall to capture food intake might fail to obtain representable data due to the varied eating habits among individuals [26, 27]. Improvements in vegetable consumption observed from Ratcliffe et al., 2011 and McAleese et al., 2007 were mainly due to the reduced intake in the control group, thus implying a small improvement in the intervention group [22, 59].

In this review, there was a small number of studies that investigated the intervention impact on children’s home F&V availability (n = 6) and consumption (n = 2). Home F&V availability is known to be a crucial determinant in affecting children’s food choices. If F&V is not readily available at home, it is difficult for the children to transfer what they have learnt from the intervention into everyday life [26, 27]. However, most of the studies reported no significant impacts of SGBP on home F&V availability and consumption. The potential explanations account for the neutral effect of SGBP in modifying children’s home F&V consumption include an increase in age (lessoning effect on parental influences), family influences, unappealing presentation of F&V in home meal, low home food security, comparatively higher cost of F&V, limited accessibility to F&V within community and media influences [73, 85, 86]. With respect to home F&V availability, the authors speculated that the failure of increasing the home F&V availability may be due to the low parental participation or response rate, which subsequently hinders the children’s F&V uptake at home, as parents remain to be the main nutrition gatekeeper [35]. Since children spend a significant amount of time at home, they are more likely to enjoy and consume food that their parents enjoy, or vice versa, and prefer foods that are readily accessible and available in the home environment. Thus, it is important to improve parents’ KAP on diet and nutrition as well as the home F&V availability and accessibility to improve children’s food choices. Besides, it may be due to other determinants such as socioeconomic status and cultural influences, as well as the accessibility and availability of such foods within the neighbourhood [87‐89].

This review reveals that parental involvement in SGBP may help to better promote children’s attitudes towards and willingness to consume F&V. Parents are known to play a fundamental role in the development and achievement of the children [96]. According to social learning theory, children learn and model from the behaviour of others through observation [97]. As children spend a substantial amount of time with parents, children’s food choice, eating behaviour and eating-related attitudes are thus hugely influenced by their parent [92, 98, 99]. Active and effective parental involvement in the SGBP, including face-to-face engagement with frequent interactions with children, provides parents more opportunities to impose positive parental modelling effects on healthy eating and enjoyment of eating F&V. Children are thus more likely to develop a preference for F&V and to make healthier food choices.

The insignificant parental effect observed in most of the studies could possibly be due to the little parental involvement element and low parents’ participation in the intervention activities. Studies conducted by Jaenke et al., 2012, Lee et al., 2017, Taylor et al., 2018, and Massarani et al., 2019 only included regular parent newsletter distribution as their main parents’ engagement activities, as newsletter could only serve as an informed purpose, parent-children’s interactions were hence limited [38, 42, 45, 54]. In addition, parents often have inflexible and overwhelming work schedules with multiple responsibilities in charge, resulting in limited time dedicated to each responsibility and hence a low parents’ participation in school activities is expected [100, 101]. Therefore, it is understandable that no significant parental effect was observed. Moreover, some of the studies did not describe the parental involvement activities in detail (e.g., the number of parents involved, and the response rate were absent), which make it difficult to assess its impacts on the effectiveness of the interventions. Future SGBP, hence, should measure the degree of parental engagement, if possible, examine what constitutes effective parental involvement and identify effective strategies to promote and maximise parental interaction with children during the intervention. To encourage more parental engagement, it requires better cooperation between the school and parents. It is suggested to schedule meetings and activities on multiple occasions to match parents’ varying schedules and be flexible in accommodating parents and families in the school programmes such as providing incentives, food or refreshments, and free transportation to minimise barriers and create an enabling environment for parents’ participation [102].

Six studies included from LMICs observed, similar to the overall findings or findings in developed countries, significant improvements in children’s preferences and attitudes towards vegetables but no significant improvement in F&V intake. Whilst improvement in food, nutrition, gardening and science-related knowledge was observed in the developed countries, it was not observed in the LMICs. The insignificant gardening effects on improving children’s knowledge and F&V consumption could be due to the inadequacy in school resources. Due to the limited number of teachers and fewer classroom materials provided, teachers from LMICs are already struggling with completing their high amount of workload. Schreinemachers et al., 2019 reported that teaching in Burkina Faso is difficult as every teacher is averagely responsible for 45 primary schoolchildren, requesting them to take on extra responsibilities and time for the implementation of an unessential gardening intervention are thus extremely hard [40]. The intervention effects might consequently be undermined as teachers are less likely to deliver the intervention programme due to the constraints of time and resources. Furthermore, the seasonal supply of vegetables also lowers the intervention effect. Due to the limited water supply in the dry season, vegetables are reported to be available for only 3 to 4 months, effect on improving children’s preference towards and intake of F&V are therefore reduced due to the limited availability and accessibility of F&V [40]. Besides, it is known that most school children from LMICs already participated in agricultural activities at home, garden-based intervention might be more appealing to those from developed countries where children lack nature experience and outdoor activities at school [40]. The parental influences on shaping children’s dietary behaviour were weak in the six studies, which is possibly due to their comparatively low education level [26, 33]. With limited knowledge and awareness of nutrition and healthy diet, parents from LMICs might be less likely to make healthier food choices for their children, family modelling impact is thus reduced.

One of the strengths of this review is that a comprehensive literature search was performed from five different databases to adequately identify most of the literature related to this topic, and potentially reduced the selection bias [103]. A robust review method was used, as two reviewers were involved to determine the inclusion and exclusion of studies independently. In addition, based on the Quality Criteria Checklist from the Academy of Nutrition and Dietetics (2016), none of the included studies was rated as high risk of bias, this demonstrates the high quality of the included studies.

There are a few limitations that cannot be ignored. The use of various measurement tools to assess the outcome measures increases the complexity of interpretation when comparing between studies. Heterogeneity of the intervention components, sample sizes, study designs and outcome measures between studies implied that the synthesis of the meta-analysis was not possible. Therefore, a single summary estimate of the impacts of SGBP failed to be generated [104]. In addition, result in this review is just a general sum up of findings from each study reported, age, sample size, ethnicity and study quality have not been weighted. Impacts of society-led or community-led gardening programmes have also not been explored and discussed, thus it is beyond the remit of this review to fully cover the impact and efficiency of society or community-led gardening programmes, a recent review by Ohly et al., 2016 provides a comprehensive overview of this topic [105]. Moreover, all SGBPs included were multifaceted with varying degrees of cooking, gardening, and nutrition components. It is thus difficult to assess which aspect of these components and how intense (or what dose) these components were most likely to be associated with positive outcomes on children’s diet and nutrition related KAP. Besides, most of the SGBP did not consider the variation of teacher experience and motivation, which are some of the determinants for programme effectiveness. The use of varying teaching approaches and enthusiasm in curriculum delivery may influence students’ learning outcomes, and potentially determine the success of the intervention [54, 106]. In addition, the review might be prone to selection bias as only studies written in English were included. Furthermore, evidence was based on studies in which the participants were predominantly US and European populations with only a small amount from Asian or other countries, therefore the findings generated may not be generalizable and transferable to the other populations.

To strengthen the impact of SGBP in promoting children’s KAP as observed from most of the successful interventions, integrated SGBP which include multiple or additional components such as nutritional education and parental involvement activities are encouraged to maximise the intervention effect. Classroom education is a crucial aspect of SGBP to effectively improve children’s nutritional-related knowledge. Thus, future SGBP is highly recommended to integrate with age-appropriate classroom education, conduct in smaller sample size with smaller child-to-staff ratio and shorter duration (~ 12 to 28 weeks), and intervene at an early age, preferably around the pre-school and primary school age, to achieve better outcomes on children’s nutritional-related knowledge and F&V consumption [107]. In addition, parental involvement in SGBP may help to promote children’s attitudes towards F&V, which may subsequently promote the intake of such foods.

Future SGBP should also adopt a multi-level approach, which covers the school, home and community environment to maximise the scope and therefore the impact of intervention. In addition, to further promote F&V consumption among children at school, it is recommended to incorporate food service into SGBP intervention. Potential ways include setting up a school salad bar in the cafeteria using the crops grown from SGBP or utilizing the produces to supplement the food in the cafeteria to increase the accessibility and availability of F&V at school. Future SGBP should also consider building relationships or partnerships with the local farmers or community gardens or, promoting or providing the students and their families the locally grown produce so as to maximise the exposure to F&V and the potential to promote such intake. Future studies should investigate the effective strategies to improve parental participation and involvement to strengthen the impact of SGBP in improving F&V consumption. In addition, it is essential for parents to acknowledge their key roles in shaping children’s eating habits. Future SGBP should offer more parental lessons and provide parents with the knowledge and tools to improve children’s eating behaviours [108]. This includes 1) offering practical advice on fostering children’s preferences towards healthier food options and increasing their willingness to consume unfamiliar food; 2) understanding the negative impact of coercive feeding practice and providing alternative options; 3) helping parents to establish a good parental role model; 4) educating the importance of not overfeeding their children and not forcing them to finish the meal when full [108]. Furthermore, more focus should be placed on investigating the long-term impact and the sustainability of the future SGBP.