THE IMPACT OF SCHOOL GARDENS ON NUTRITION OUTCOMES IN LOW-INCOME COUNTRIES

Introduction

Low quality diets have significant human health risks (Willett et al., 2019). Of the world’s population of 7.2 billion, more than 2.1 billion people are overweight or obese and 2 billion people are micronutrient deficient. Hence there is a growing need to transform current food systems to ensure healthy diets. As part of this, there is a need for good policies and programmes that promote healthy eating habits.

This requires a sound understanding of drivers of eating habits. It is known that taste preferences and food attitudes are important drivers alongside lifestyles, convenience, affordability, nutrition awareness, and policies (Kearney, 2010; Traill et al., 2014). Taste preferences and food attitudes are established early in life and tend to persist through to adulthood (Kelder et al., 1994; Birch, 1999; Cooke, 2007). This understanding has stimulated interest in interventions that nudge children to adopt healthy eating habits. School garden programmes are one such type of intervention.

School garden programmes have become popular for improving food, nutrition, and health behaviour of schoolchildren in both high- and low-income countries (FAO, 2005; Ozer, 2007; Parmer et al., 2009; Christian et al., 2014a; Hutchinson et al., 2015; Triador et al., 2015; Nury et al., 2017). Such programmes usually bundle hands-on practical education in school gardening with nutrition education. School gardening training is expected to give children the knowledge and the ability to produce food, while nutrition education promotes children’s demand for quality food. This combination of supply- and demand-side interventions is a key feature of school garden programmes.

The use of school gardens to promote healthy eating behaviour in low-income countries is conceptually appealing because of high rates of child malnutrition. As a result, there has been a substantial increase in such programmes in low-income countries. However, the evidence basis for their impact on food behaviour in low- income countries is very limited as there are no impact evaluations in the literature.

This chapter describes the results of three recent impact evaluations, each using a cluster randomised controlled trial design, of school garden programmes in Bhutan, Nepal, and Burkina Faso with the objective to assess the current evidence for the impact of school garden programmes in low-income countries. The three studies that this chapter is based on are the largest evaluation of school garden programmes to date.

The results show that school garden interventions significantly improved children’s knowledge about food and agriculture in all three countries. For Bhutan and Nepal, the results showed improvements in children’s awareness of fruit and vegetables and their stated preferences for eating them. In none of the countries did these changes translate into clear increases in fruit and vegetable consumption. The results therefore suggest that for school garden programmes to contribute to healthier food choices a more holistic intervention design may be needed that addresses the household-level supply of healthy foods and influences the food choices of parents, considering affordability and availability factors.

The chapter starts by reviewing the existing evidence for school garden programmes. It then presents the results of three recent studies from Bhutan, Burkina Faso, and Nepal. The chapter concludes with some recommendations on how to better design school garden programmes aimed at influencing food choices in low-income countries.

Current evidence for school garden programmes

School garden programmes need to be designed in response to the food environment and lifestyle of people in the target location. In high-income countries, school gardens often try to promote greater physical activity levels among children, encourage social interaction, promote healthier eating habits, and give children a better understanding of food production while trying to instil a love for nature. Such goals can also be found in lower income countries, but the emphasis is usually more on building children’s resilience by giving them the ability to grow food in a sustainable way through the transfer of agricultural knowledge and skills, and increasing the consumption of healthy foods such as micronutrient-dense fruit and vegetables.

Virtually all the evidence for the impact of school gardens has come from higher-income countries (Ozer, 2007; Lineberger and Zajicek, 2009; Parmer et al., 2009; Morgan et al., 2010; Ratcliffe et al., 2011; Christian et al., 2014b), while there was no evidence for lower-income countries until this project was completed. In the following we will describe this evidence based on four review papers before turning our attention to lower-income countries.

Blair (2009) reviewed 12 quantitative impact studies of school garden programmes in the United States. All 12 quantitative studies collected pre- and post-intervention data, but none used an experimental design in which schools would be randomly assigned to a treatment or control group. These 12 studies used samples of 1 to 5 schools and no more than 500 schoolchildren. The review showed that 9 out of the 12 studies showed positive outcomes in the area of science achievement, but only one study showed a significant increase in fruit and vegetable consumption that could be attributed to the school garden programme.

Robinson-O’Brien et al. (2009) reviewed 11 studies for the United States that examined the impact of garden-based nutrition education programmes on preferences, willingness to taste, and intake of fruit and vegetables. Of these 11 studies, four reported effects on fruit and vegetable intake and three of these reported a significant increase. Six studies reported results on fruit and vegetable preferences, and two of these reported a significant increase. Three studies reported results on willingness to taste fruit or vegetables, and two of these reported a significant increase. The authors therefore concluded that, based on the limited evidence available, garden-based nutrition-education programmes may have the potential to lead to improvements in fruit and vegetable intake, willingness to taste, and preferences among students.

Langellotto and Gupta (2012) specifically reviewed the evidence of garden-based nutrition education programmes on student’s nutrition knowledge, preferences for fruit and vegetables, and consumption of fruit and vegetables. They reviewed 20 studies from the United States and found that nutrition education programmes that included a gardening component were more effective at increasing vegetable consumption in children than nutrition education programmes without such gardening component.

Ohly et al. (2016) conducted a systematic review of quantitative and qualitative evidence of school gardening programmes. The review identified 21 quantitative studies, 16 qualitative studies, and 3 mixed methods studies. The 40 studies all came from high-income countries (United States, United Kingdom, Australia, and Portugal). Thirteen studies reported the effect on healthier food preferences and eight found a significant effect. Ten studies reported effects on food knowledge and attitudes, and seven of these reported a significant increase. Thirteen studies reported the effect on fruit and vegetable consumption, and only two found a significant increase. Hence, this review shows that school garden programmes are more capable of influencing children’s food preferences, knowledge, and attitudes but less able to influence actual food behaviour in terms of a significant increase in children’s fruit and vegetable consumption.

These reviews for high-income countries point out that there is no strong evidence that school garden programmes improve children’s consumption of fruit and vegetables. School garden programmes are more likely to improve science scores, food knowledge, and food preferences than they are able to influence actual behaviour towards healthier food choices.

Results of three recent experimental studies for low-income countries

To fill the evidence gap for the impact of school garden programmes in low-income counties, the Swiss Agency for Development and Cooperation (SDC) funded the project ‘Vegetables Go to School: Improving Nutrition by

Agricultural Diversification’ (2013-2017). Through this project, the World Vegetable Center worked with national partners in Burkina Faso, Tanzania, Indonesia, Bhutan, Nepal, and the Philippines to design and implement school garden programmes. Quantitative evidence for impact was generated for three of these countries - Bhutan, Nepal, and Burkina Faso. Results have been published previously (Schreinemachers et al., 2017a; Schreinemachers et al., 2017b; Schreinemachers et al., 2019) and the purpose of this chapter is to synthesise and highlight the evidence from these studies.

Research design and data collection

Each study used a cluster randomised controlled trial design. Treatment and control schools were selected randomly from a larger list of primary schools in a particular region of the country. For each school, baseline data were collected at the start of the school year, next the intervention was implemented, and end line data were collected at the end of the school year for the same students (Table 8.1). In Nepal and Burkina Faso the study was repeated to increase the robustness of the results. For these two countries, control schools in year-1 became treatment schools in year-2. In Bhutan the data collection started only in the second year of the project.

TABLE 8.1 Data collected for the study

Bhutan

Nepal

Burkina Faso

C

T

A

C

T

A

C

T

A

Year-1:

# schools

-

-

-

20

10

30

10

10

20

# students, baseline

904

466

1,370

500

500

1,000

# students, endline

-

-

-

882

454

1,336

499

501

1,000

# students, total[1]

-

-

-

846

429

1,275

491

488

979

Year-2:

# schools

9

9

18

10

10

20

10

10

20

# students, baseline

265

260

525

433

394

827

400

400

800

# students, endline

258

259

517

385

428

813

400

400

800

# students, total [1]

235

233

468

416

369

785

389

392

781

Source: (Schreinemachers et al., 2017a; Schreinemachers et al., 2017b; Schreinemachers et al., 2019). Notes: C = Control; T = Treatment; A = Sum of Control and Treatment.

Outcome indicators were selected from a review of previous impact studies of school garden programmes. The selected outcome indicators represented stepwise changes in the intervention’s impact pathway from increased awareness about fruit and vegetables to improved knowledge of food, nutrition, and WASH (water, sanitation, and hygiene) as well as knowledge of sustainable agriculture and increased preferences for healthy foods and food behaviour in terms of increased fruit and vegetable consumption. Details about variable definitions and measurement can be found in the published papers. Data were collected using relatively simple questionnaires with photos and multiple-choice questions to aid children’s comprehension. The text was brief, and the phrasing was simple. The questionnaire was designed in English and translated into the local language.

Intervention designs

The school garden programmes in each country were designed in collaboration between the Ministry of Agriculture and the Ministry of Education. For details see (Bhattarai et al., 2016; Bhattarai et al., 2017; Ouedraogo et ah, 2017; Rai et ah, 2017). This collaboration across ministries was an important aspect of the project. A nutrition expert from the Ministry of Health in Burkina Faso and Nepal and from a local university in Bhutan also contributed to the project. Each programme design included three intervention components that were concurrently implemented in all treatment schools and targeted children in elementary education:

School garden component: School gardens were established in each treatment school for the cultivation of about eight different vegetables. Emphasis was given to vegetables that were culturally accepted, nutrient-dense, and easy to grow. The school garden designs generally included raised planting beds, fences to prevent livestock from entering the garden. Irrigation water tanks were installed where water availability was a constraint. Other improvements were made depending on the local situation.

Education component: Complementary education was provided about agriculture, nutrition, and hygiene. A curriculum was developed for 23 weeks and teachers received training on how to use it. Topics covered were food groups, the health benefits of vegetables, food and body hygiene, and school and environmental sanitation. The school garden and its products were used by teachers to illustrate mathematics, biology, moral education, and language. The agricultural education emphasised learning by doing and was regularly conducted in the school garden.

Involvement of parents, local fanners and other community members: Parents helped preparing the school garden and fencing it with locally available materials. Depending on the local situation, vegetables were supplied to the school canteen or given to the children to bring home. In many instances, local farmers helped with making nurseries and preparing seed beds and local extension officers visited the schools to provide technical advice as needed.

The three intervention components are jointly expected to raise children’s awareness about vegetables and increase their knowledge. This increased knowledge is then expected to stimulate their preferences for healthier food choices including vegetables, which should then lead to better food choices and improved nutrition status.

Results

Table 8.2 shows the average treatment effects, which is the difference in mean outcomes between students assigned to the treatment and students assigned to the control. For means and standard errors we refer to the original papers.

TABLE 8.2 Impact of the school garden intervention on nutrition outcomes in Bhutan, Nepal, and Burkina Faso, average treatment effects showing marginal effects at means

Outcome variable

Bhutan

Nepal

Burkina Faso

Year-2

Year-1

Year-2

Year-1

Year-2

Awareness:

% of fruit and vegetables correctly named

17.9 ★ ★★

29.4 ★ ★★

12.8 ★ ★★

3.3

2.5

Knowledge:

% of correct answers on sustainable agriculture

15.2 ★ ★

21.7 ★ ★★

16.7 ★ ★★

4.5

5.7

% of correct answers on food, nutrition, and WASH

-5.2

13.8 ★ ★★

14.6 ★ ★★

6.1 ★ ★

7.7 ★ ★

Preferences:

% of fruit and vegetables liked

9.5 ★ ★

15.8 ★ ★★

19.1 ★ ★★

-1.4

12.3

Behaviour:

% of children that ate vegetables

11.7 ★ ★

2.35

0.91

20.2

2.7

% of children that ate fruit

-3.6

-0.75

7.11

-6.5

-3.9

# of different vegetables eaten

0.19

0.09

-0.04

0.3

0.2

# of different fruits eaten

-0.09

0.07

0.08

NA

0.1

Students (n)

468

1,275

785

979

781

Sourer. (Schreinemachers et al., 2017a; Schreinemachers et al., 2017b; Schreinemachers et al.. Under review).

Notes: ***/> < 0.01, **p < 0.05, * p < 0.10. NA = Not available (the average treatment effect could not be estimated as the number of different fruits eaten in the sample of students in Burkina Faso in year-1 was near zero).

The results show that the school garden programme, which was similarly designed and implemented in each country, had more significant effects on nutrition outcomes in Bhutan and Nepal than in Burkina Faso. For the latter country, only one of the eight outcome variables showed a significant increase.

For Bhutan and Nepal, the school garden programme increased students’ awareness of fruit and vegetables. For Nepal and Burkina Faso, the school garden programme increased students’ knowledge of food, nutrition, and WASH. For Bhutan and Nepal, but not for Burkina Faso, the school garden programme also enhanced students’ knowledge of sustainable agriculture. It is, however, noted that for Burkina Faso, when pooling the two years together, the effect on agricultural knowledge was also significant (p < 0.05).

For Bhutan and Nepal, but not for Burkina Faso, the school garden programme promoted children’s preferences for eating fruit and vegetables. Only in Bhutan, did the school garden programme increase the percentage of children eating vegetables. The effect of the school garden programme on fruit consumption or the number of different fruits or vegetables consumed was insignificant for all years and all countries.

These studies therefore broadly show that school garden programmes in low-income countries can be effective in increasing students’ awareness of fruit and vegetables, their knowledge of food and agriculture, and their preferences for eating fruit and vegetables, but school garden programmes appear less effective in changing students’ actual food behaviour towards the increased consumption of fruit and vegetables, at least in the short-term.

  • [1] Students outside an age range were dropped from the sample with the age range being 9—15 yearsold for Bhutan, 10—15 years old for Nepal, and 8-14 years old for Burkina Faso.
  • [2] Students outside an age range were dropped from the sample with the age range being 9—15 yearsold for Bhutan, 10—15 years old for Nepal, and 8-14 years old for Burkina Faso.
 
Source
< Prev   CONTENTS   Source   Next >