This study was conducted with the approval of the Institutional Review Board of K University (KYU2022-3-16-001). The research subjects who were users of a facility voluntarily participated in this study. After receiving their application through the facility, we contacted them individually and obtained their consent to participate in the study. The specific details of this research procedure are shown in Fig. 1.

The experimental and control groups were assigned using simple random sampling. A pre- and post-test of the experiment was conducted by the researchers and a professor of rehabilitation medicine. The evaluation tools include the Korean-Cognitive Impairment Screening Test, the Short Form Geriatric Depression Scale, and the Senior Fitness Test. In addition, the experimental group’s preferred activities for each area were surveyed after implementing the program. The agro-healing program was conducted once a week over a total of 12 sessions from May 18 to August 10, 2022 at the Y Senior Welfare Center in Jeonju, Jeollabuk-do. The time required for the program was 120 minutes per day, and the program proceeded as follows: 10 minutes was spent on an introduction and preparations, followed by a 50-minute outdoor agro-healing program (total of 60 minutes), followed by a 10-minute break, and then a 50-minute indoor agro-healing program.

The subjects of this study were 19 users of the Y Senior Welfare Center in Jeonju-si, Jeollabuk-do, who understood the purpose of this study and agreed to participate in the survey, including 10 subjects who participated in the program, and 9 subjects who did not participate in the program. Selected subjects were seniors aged 65 years or older who were able walk independently, and individuals who had been diagnosed with diabetes, were taking related medications, or had allergic diseases caused by weeds or pollen were excluded. As for the general characteristics of the subjects, in terms of gender, there were 5 males (55.56%) in the control group and 3 (30.00%) in the experimental group. The average age was 68.4 years for the control group and 76.5 years for the experimental group (Table 1). The control group consisted of 3 people between 65 and 69 (33.34%), 2 people each in the age groups of 75 to 79, 80 to 84, and 85 to 89 (22.22%), and 2 people whose exact ages were not disclosed (22.22%). The largest age groups in the experimental group were 75 to 79 years old and 80 to 84 years old, with 4 people each (40.00%), followed by 70 and 74 with two people (20.00%). In terms of education level, the control group included 3 high school graduates (33.34%), 2 university graduates (22.22%), and 1 person each in the no education, elementary school graduate, and middle school graduate categories (11.11%); whereas the experimental group consisted of 6 high school graduates (60.00%), 2 elementary school graduates (20.00%), 1 middle school graduate (10.00%), and 1 person who did not disclose their education level (10.00%). There appeared to be no differences in general characteristics between the control and experimental groups.

Stages and areas of the program based on cognitive-based interventions for the elderly (Jo, 2017) were organized, and then program topics were presented. The program was then established by modifying and supplementing the type, intensity, and evaluation method of activities in consultation with one professor of horticulture, one professor of rehabilitation medicine, two social workers from Y Senior Welfare Center, and five researchers from the National Institute of Horticultural and Herbal Science (NIHHS) under the Rural Development Administration. The agro-healing program for each session consisted of outdoor gardening conducted in a vegetable garden, followed by indoor gardening using garden plants and making craft products.

Pre-measurement of the assessment tools in this study was performed twice, in the first and second weeks of May, and post-measurement was performed twice, in the third and fourth weeks of August. The assessors were a professor of rehabilitation medicine and five researchers from the NIHHS who visited the Center and measured the assessment tools in the office and activity room on the first floor. One researcher determined and confirmed the measurement process and order for each area of the subjects. For the Cognitive Impairment Screening Test, two researchers asked questions and recorded the answers. For the Short Form Geriatric Depression Scale, one researcher asked questions and recorded the answers. For the Senior Fitness Test, a professor of rehabilitation medicine measured the subjects’ physical fitness using the tool, and one researcher recorded it. Meanwhile, for the post-measurement of program preference, program contents and photos were attached to the office wall, and subjects were asked to check the program and attach stickers to their preferred activities.

In this study, a statistical analysis was performed using SAS 9.4 (SAS Institute Inc., Cary, NC). The general characteristics and activity preferences of the subjects were analyzed using descriptive statistics. Using Fisher’s exact test, the homogeneity of general characteristics of the control and experimental groups was assessed before carrying out the agro-healing program. Wilcoxon’s rank-sum test was performed to determine the difference in the results between the control and experimental groups before and after the agro-healing program, while Wilcoxon’s signed-rank test was conducted to determine the difference in the results within each group.

The comparison of the average scores of sub-items (orientation, memory, attention, visuospatial function, language function, and executive function) for cognitive functions of the control and experimental groups did not show significant results, proving homogeneity between the groups (Table 3). Comparing the mean scores on cognitive function in the pre- and post-test, among the control group, the difference in scores was not statistically significant: 4.3 and 4.6 points for orientation, 2.2 and 2.7 for attention, 3.7 and 4.2 for executive function, 6.9 and 7.7 for memory, and 6.9 and 7.7 for language function, respectively (Table 3). For the experimental group, the mean scores of the pre- and post-test were 2.2 and 2.6 points for attention, 3.5 and 4.1 for executive function, and 7.2 and 8.7 for memory, respectively, indicating that the difference in scores was not statistically significant. This result cannot support the finding that horticultural therapy had a positive effect on the cognitive function of the elderly with mild cognitive disabilities (Joo, 2023) and helped those with dementia improve their cognitive function (Lee, 2021). It seems that this result was affected by the subjects’ condition on the day of measurement and the limitations of controlling exogenous variables occurring to them. Therefore, it is necessary to verify the effectiveness on cognitive function reflecting changes in lifestyle, physical, social, and emotional aspects that affect changes in cognitive function of the elderly (Yoon, 2022).

Meanwhile, changes in the subjects’ cognitive aspects were shown through the researchers’ subjective observations of each session. At the beginning of the program, the subjects had difficulty remembering plant names and characteristics, counting leaves and flowers, and answering questions about cultivation methods, but in the second half, appeared to be able to perform such tasks or answer questions comfortably. This seems to be because the changes were made through memory training (Son et al., 2010), which is a process of repeated questions and answers after observing changes in plants in each session. In addition, the subjects appeared to feel burdened at the beginning of gardening activities using plants, but as the sessions progressed, it was observed that they became curious about new activities (Kim, 2008), focused on task descriptions, and improved their task performance (Kim and Yu, 2013).

The comparison of the mean depression scores of the control and experimental groups did not show significant differences, proving homogeneity between the groups (Table 4). In the pre- and post-test, the mean depression score of the subjects was 2.4 and 2.0 points for the control group, and 4.0 and 2.8 for the experimental group, respectively, indicating that the difference in scores was not statistically significant. This result cannot support the findings that gardening programs for cognitive rehabilitation reduced depression among the elderly (Lee, 2008; Lee, 2021). Although there were subjects who showed changes in the mean score, this result seems to have been affected by the difficulty in obtaining significant results due to the small number of research subjects. Therefore, in evaluating the depression of subjects, not only quantitative but also qualitative assessment is required, along with a higher number of subjects (Kim, 2018), and physical and family support factors that affect depression of the elderly (Oh, 2014) should also be considered. Meanwhile, looking at the change in the subjects’ emotions through the researchers’ subjective observations for each session, the subjects who passively participated in the beginning of the program appeared to feel a sense of accomplishment in the outcomes obtained from gardening work (Lee, 2008), and gave feedback on their activities and actively participated in the latter half of the program. Moreover, it was observed that the subjects expressed positive emotions while observing the growth of plants and harvesting in the garden (Cho, 2022).

The comparison of the mean scores of sub-items for physical function and fitness of the control and experimental groups did not show significant differences, proving the homogeneity between the groups (Table 5). The control group showed no significance difference in the mean scores: 15.6 and 17.4 times, respectively, in the 30-second chair stand test of lower body strength in the pre- and post-test; 19.3 and 23.2 times in the 30-minute dumbbell workout test of upper body strength; 102.3 and 108.8 times for 2-minute step in place test of cardiopulmonary endurance; 2.5 and 3.4 cm in the seated toe touch test of lower body flexibility; and 10.3 and 8.8 seconds for an 8 ft round-trip walk test of agility and dynamic balance. The experimental group also showed no statistically significant difference: 14.4 and 18.6 times, respectively, for the 30-second chair stand test of lower body strength in the pre- and post-test; 20.7 and 21.9 times for the 30-minute dumbbell workout test of upper body strength; 101.2 and 113.4 times for the 2-minute step in place test of cardiopulmonary endurance; 2.6 and 10.8 cm for the reaching hands behind the back test of upper body flexibility; and 7.8 and 7.3 seconds for an 8 ft round-trip walk test of agility and dynamic balance. This result cannot support the finding that the physical function and fitness of the elderly were improved through horticultural therapy (Han, 2018). In this study, the program was conducted in one session per week. Considering that exercise three times a week is needed to improve physical function through activities (Lee, 2014), in future studies, it seems necessary to reduce the interval between sessions and increase the number of sessions per week to achieve physical fitness. Meanwhile, examining the subjects’ physical activity through the researchers’ subjective observation for each session, it seems that the repetitive tasks of gardening, including watering, weeding, picking leaves, cutting stems, and harvesting, improve the elderly’s muscular strength (Son et al., 2010), and that exercises that improve hand function (Kang, 2002) can help elderly people perform tasks in the area of physical fitness as their target.

After the 12-session agro-healing program, a preference survey for each area with multiple selection options was conducted for the experimental group. The survey results are shown in Table 6. First, when the subjects were asked about activities that required a lot of physical exercise, the most common answer was creating (flower) beds (5 people; 27.78%), followed by playing herb-ball games (4; 22.22%), planting seedlings (3; 16.66%), and making garden-plant wreaths (2; 11.11%). This suggests that various physical activities were carried out in the process of creating an outdoor flower bed (Han, 2018). Second, when they were asked about activities that required a lot of mental work, the most common answer was playing games with herb balls (5 people; 27.78%), followed by making garden-plant wreaths (4; 22.22%) and decorating vases with twigs (3; 16.66%), and checking their healthy lifestyle (2; 11.11%). This indicates that the process and steps of playing games with herb-balls, and the craft work that required following instructions for making herb-balls could stimulate their cognition (Lee et al., 2000). Meanwhile, when they were asked about the activity that made them feel most at ease, the most common answer was expressing their feelings with pressed flowers (3 people; 16.66%), followed by hand exercises for cognitive health, herb-ball games, herbal foot baths, and farm parties (2 each; 11.11%). They answered that the activities of expressing their feelings using pressed flowers, expressing the characteristics of plants, relaxing the body, and sharing the harvest helped their emotions. Of the 12 activities, the activity with the highest preference was playing herb-ball games (11 people; 20.39%), followed by making wreaths with garden plants (6; 11.11%), and creating flower beds and decorating vases with twigs (5 each; 9.26%). Based on this, it seems that the activities of the elderly in agro-healing programs should provide interest and fun through games using scents and such, and tasks that produce outcomes using plants rather than familiar and simple ones (Kim and Yu, 2013). Considering that the initial stage of gardening activities, which involves creating flower beds and planting plants, was preferred, it appears to be necessary to involve subjects in future agro-healing programs from the beginning of planning gardening activities, and to select crops that can be continuously harvested and used in the intermediate stage for various activities.