The purpose of this study and experimental design are approved by the Institutional Review Board of G University (Approval No.: GIRB-A19-Y-0056). The announcement for participant recruitment was posted on the bulletin board of a health center in H county, and participants were recruited from August 9 to September 1, 2019 at the health center located in H county, Gyeongsangnam-do. The participants were caregivers taking care of the elderly with dementia who can participate in both pretest and posttest surveys and are willing to participate in eight sessions of the program. The researchers verbally explained to the selected participants about the purpose and process of research as well as participant guidelines during the pretest according to the bioethical regulations, and the research was conducted after obtaining consent for participation. 30 ultimately selected participants were randomly sampled and divided into two groups, 15 in the control group and 15 in the experimental group, and the experiment was conducted in a pretest-posttest quasi-experimental design. The same program was provided for the control group as a reward for participation after the experimental group was done with the entire program.

The horticultural therapy program was carried out twice a week from September 2 to 25, 2019 in total eight sessions. The purpose of the program was to alleviate depression and caregiver burden and improve QOL for caregivers of the elderly with dementia. Detailed goals were to understand dementia, establish a positive self-image, relieve stress, and rebuild family relations and relationship with the person with dementia. These goals were set up with reference to those applied in psychoeducation, counseling and psychotherapy, and mindfulness-based psychotherapeutic intervention programs that have been used to solve psychological and emotional problems of caregivers of the elderly with dementia. According to the findings from other intervention programs, the programs that have the goals to understand dementia, establish a positive self-image, determine the cause of stress faced in the caregiving process and learn how to resolve this, and find the meaning of family relations or relationship with the person with dementia are helpful in alleviating depression and caregiver burden and improving the QOL for caregivers of the elderly with dementia (Cheng et al., 2019; Cheong et al., 2020; J.Y. Kim et al., 2016; Kuo et al., 2016; Livingston et al., 2013; Seo et al., 2018; Yang et al., 2019; Waelde et al., 2017). Based on the purpose of such programs, the program used in this study was organized with consultation of one professor of horticultural therapy, one professor of psychiatry, one professor of social welfare, and one PhD candidate in horticultural therapy (Table 1). The program was carried out in two steps, personal crisis intervention (Sessions 1–5) and interpersonal crisis intervention (Sessions 6–8), as suggested by Livingston et al. (2013), Kuo et al. (2016), and Yang et al. (2019) to solve the major problems faced by caregivers caring for dementia patients. Moreover, a series of activities were formed by comparing them to the process of plant cultivation, such as ‘Garden Formation (Session 1)’→ ‘Sowing (Session 2)’→ ‘Growth (Session 3)’ → ‘Flowering (Sessions 4–5)’ → ‘Fruition and Harvest (Sessions 6–8)’.

More specifically, Session 1 ‘Garden Formation’ was ‘Making a dish garden using moss’ with the topic ‘Empathic garden called dementia’. Session 1 is the introductory step in which the participants brainstormed about dementia, which is a shared topic of conversation for them, with keywords such as ‘what comes across the mind regarding dementia’ and ‘hardships from caring for dementia patients’. The process of forming consensus and rapport about dementia with the participants by sharing one another’s thoughts was reinterpreted in terms of ‘Forming an empathic garden’. After that, the participants wrapped up by expressing a moss garden as their own rest area to relieve stress from caregiving in the dish garden.

Session 2 ‘Sowing’ had the topic “A flower called ‘I’”, comprised of activities with detailed goals of improving positive self-esteem of individuals and future-oriented thinking through ‘Sowing garden nasturtium seeds’ in which the participants wrote down their wishes on flower pots and sow seeds and ‘Making a name tag using a pressed flower’ in which they introduce themselves using adjectives that express their strengths like the plant name tags in the garden. All the attention of caregivers of the elderly with dementia is focused on the dementia patients, which makes them spend most of their personal time caring for the patients, face difficulties in leisure or social activities, and even give up on their everyday life. In the process, the increased caregiver burden causes them to lose their individuality and, in extreme cases, even leads to suicidal ideation or attempt (J.Y. Kim et al., 2016). As such, to resolve the issue of losing oneself due to caregiver burden, the participants sowed seeds while thinking of them as their wishes in Session 2, determined their strengths and introduced them, through which they could form positive future-oriented thinking and sense of self.

In Session 3 ‘Growth’, the participants made terrariums with the topic “Until the flower called ‘I’ blooms”. A terrarium is to create a cyclical ecosystem by putting soil and plants inside a container. In this session, the participants were to associate the positive aspects of their attributes (individual character, memories, etc.) with various materials put into the terrarium in the process of embodying the terrarium into themselves. They used various colors of sand for soil that lay the groundwork for plants to live, thereby expressing their personalities or memories related to each color. For plants, they looked at the plants and tried to find the resemblance to them and express the planted plant as themselves. This process enabled them to explore what they are made of to the point of what they are today and objectively express and see themselves through the terrarium.

Sessions 4 and 5 ‘Flowering’ were carried out with the topic of flowers and scents. Moreover, in each session, activities to improve self-esteem were carried out as a wrap-up of personal crisis intervention, such as having the participants made and give a present for themselves as they have lost themselves by focusing only on dementia patients in the caregiving process. Session 4, with the topic ‘Bursting into full bloom’, encouraged the participants to think of an image of flowers in full bloom, having them talk about their memories of full-blown flowers and going outside to enjoy flowers in spring. They associated the full-blown flowers with the golden days of their lives, and brainstormed on the keyword ‘Spring of my life’. After that, they compared scandia moss with cherry blossoms, created a model of full-blown cherry tree and visually embodied their personal image of ‘Spring of my life’. Moreover, for making a gift for oneself, there was ‘Making scented soap’ in which the participants mixed materials that are right for their skin type and mold them into any shape they want. These activities were to help them recall their most positive memories, objectify and view them, thereby forming a positive self-image that was set as a detailed goal.

In Session 5 related to the scent of flowers, the participants were to make a gift for themselves with the topic ‘My own scent’ and carried out ‘Making aromatic roll-on perfume’ and ‘Making natural skin toner’. According to J.Y. Kim et al. (2016), there must be counseling and programs to promote mental health including immediate intervention and post management of caregiver burden and stress to manage personal crises such as suicide of caregivers of the elderly with dementia. Accordingly, the detailed goal of Session 5 is to learn how to face and manage stress, as the last step of personal crisis intervention. To achieve this goal, the participants faced their individual stress through brainstorming of their current stress, the situation in which they are stressed, and how they each relieve their stress with the keyword ‘stress’. Then they were to smell the aroma of lavender, grapefruit and rose wood essential oil that are known to help relieve stress by calming the nerves (Bikmoradi et al., 2015; de Almeida et al., 2009; Tanida et al., 2005). After that they shared how they feel about each scent, and made roll-on perfume by mixing essential oils. Then, with the method of using roll-on perfume they had made, they were encouraged to perceive stress when they face a stressful situation and conduct behavior that can alleviate stress (acupressure using roll-on perfume), teaching them the cognitive behavioral method to relieve stress in a different way. After that, in the activity of making a gift for oneself, the participants chose natural ingredients that are suitable for individual skin type and made skin toner. A gift presentation ceremony was held, in which the therapist complimented the participants on their individual efforts to care for the elderly with dementia and presented the gifts. This aromatic roll-on perfume and natural skin toner making process was intended to make individuals face their stress, learn how to manage it and improve self-esteem.

Sessions 6–8 are the conclusion part of the program with the meaning of ‘Fruition and Harvest’. Moreover, these sessions are the interpersonal crisis intervention step, in which detailed goals were established to resolve issues mentioned in studies on caregivers of the elderly with dementia (J.Y. Kim et al., 2016; Kuo et al., 2016; Livingston et al., 2013; Seo et al., 2018; Yang et al., 2019), such as interpersonal problems in caregiving like deterioration of family relations, dysfunctional communication with the elderly with dementia, and frustrations about social life and relationships.

In Session 6 ‘Separate yet together’, the participants carried out activities on family relations. If there is an elder with dementia within the family, the family experiences change in the new family system among the members due to the matter of caregiving for the elderly with dementia. Here, if the family relations face a negative change, this leads to family conflicts that result in decreased psychological and emotional support resources of caregivers and interpersonal problems, which add to the caregiver burden. On the other hand, suitable changes in the family system operate the support function of the family, which brings positive effects that not only solve interpersonal problems but also reduce the caregiver burden (Seo et al., 2018). Therefore, interventions for caregivers of the elderly with dementia often deal with resolution of dysfunctional family relations. Accordingly, the ‘Herbarium making’ activity was carried out with the detailed goal of reestablishing the meaning of family relations destroyed in the process of caring for the elderly with dementia. In the introductory step, the participants drew a family tree to examine their relationship with each family member, and explored the meaning of family in this process. Then, in the development step, they compared the family members to various plant materials such as dried and preserved flowers prepared, and filled the herbarium container with these materials, expressing the image of family from the view of caregivers caring for the elderly with dementia in the herbarium. This process helped the participants reestablish the meaning of relationship with each family member.

In Session 7 ‘My companion’, the participants faced the conflicts with the elderly with dementia and participated in ‘Making a wreath’ with the detailed goal of reorganizing relationships. Caregivers taking care of the elderly with dementia sometimes take a dysfunctional measure such as avoiding their problems as a coping strategy in their relationship with the dementia patient. This avoidance-focused coping strategy develops burnout and negative emotions of caregivers, whereas the task-focused coping strategy that helps the caregivers actively face, explore and improve such relationship is known to reduce negative emotions and caregiver burden (Iavarone et al., 2014). Therefore, interventions for caregivers require activities that face the relationship with the dementia patient. Accordingly, this activity helped the participants compare each stem of Cornus walteri with their good and bad memories of their relationship with the patient, through which they could face the relationship building between the participant and the elderly with dementia. Later, to turn the negative aspects into positive aspects of their relationship with the elderly with dementia, the participants decorated the wreath frame made of Cornus walteri with tillandsia that is mentioned as an air purifying plant by the media (Rural Development Administration, 2016; Yun, 2018), thereby making a dreamcatcher that has the meaning of giving sweet dreams by controlling nightmares (Dreamcatcher, n.d.). The participants face their relationship with the dementia patients through the process of making a wreath, and eliminated their negative relationships in the process of giving meaning to the wreath with objects (tillandsia and dreamcatcher) that reduce negative things like particulate matter or nightmares, thereby switching the meaning of relationships into a positive direction.

Finally, in Session 8 ‘Bearing fruit’, the participants carried out activities such as ‘Making a yearbook’ and ‘Making and giving herbal tea bags’ with the detailed goals of ending the program and improving social support of self-help groups. To begin with, the participants made a yearbook using photos from their activities and pressed flowers as they reflect on the activities in the last seven sessions. Then they sampled various herbal teas and mixed the ones they prefer to create their own herbal tea bags. They exchanged them as gifts with other participants and ended the program. Caregivers of the elderly with dementia tend to experience social activity constraints due to caregiving and interpersonal problems, and dementia family programs are carried out to solve these problems. Dementia family programs promote cognitive information about dementia, improve coping skills, and resolve psychological and interpersonal issues such as caregiver burden, depression and stress by providing psychoeducation, counseling or self-help groups. The self-help groups formed in such programs help caregivers solve interpersonal problems and increase voluntary support capacity by building a social network based on bonds among the caregivers (Park and Park, 2015; Seo et al., 2018). Therefore, the ending of this program is not an ending of all relationships but formation of self-help groups, reflecting on the activities they have done together and improving social support and relationships in the process of exchanging gifts they made themselves.

The program was carried out by one main host (Korean Horticultural Therapy and Well-being Association Horticultural Therapist Registered-Grade 2) and one assistant. Each session was from 90 to 120 minutes including 10 minutes of break time, and the program was carried out in the multipurpose program room in the health center of Haman-gun.

For survey items, this study collected demographic characteristics such as gender, age, and period of providing care for the elderly with dementia, using scales such as depression, life quality and caregiver burden of caring for the elderly with dementia through the horticultural therapy program.

To analyze the results after the horticultural therapy program, this study excluded the participants that did not participate in all sessions of the program or did not conduct the posttest. Accordingly, the analysis was conducted using pretest-posttest data of 19 participants (10 in the experimental group, nine in the control group) out of 30 participants that were initially recruited. The analysis was conducted using IBM SPSS Statistics 25.0 Program. As a result of conducting the Shapiro-Wilk test of normality for the pretest results to determine the data analysis method, all items did not have normality. Accordingly, along with basic descriptive statistics, the pretest-posttest between-group difference through the Mann-Whitney U test as a non-parametric method and the pretest-posttest within-group difference through the Wilcoxon signed rank test were tested at the 95% significance level.

All the participants were female, and their average age was 58.78 in the experimental group (SD = 9.56), and 61.3 in the control group (SD = 12.40). The average caregiving period of the participants was 6.00 years in the experimental group (SD = 3.12), and 4.00 years in the control group (SD = 2.40). Moreover, there was no statistically significant difference found as a result of conducting a statistical analysis to test the homogeneity on the average age and average caregiving period of the experimental group and control group.

Table 2 shows the result of conducting the preliminary test of homogeneity on scores of depression, QOL and caregiver burden. The analysis result showed that there is no significant difference between groups. Accordingly, the two groups are homogeneous in terms of depression, QOL and caregiver burden.

We tested the posttest mean difference of depression between the groups after the horticultural therapy program and analyzed the pretest-posttest mean by group (Table 3). As a result of testing the posttest mean difference between the groups, there was no significant difference between the groups in the total score of depression and detailed items.

In the control group, the detailed items increased in the posttest, even though they were not statistically significant: depressive emotions increased from 10.33 (SD = 3.54) to 12.22 (SD = 3.83), positive emotions (reverse scoring) from 9.78 (SD = 1.20) to 9.89 (SD = 1.54), physical inactivity from 10.33 (SD = 2.65) to 11.78 (SD = 1.86), and interpersonal relationship from 2.56 (SD = 0.88) to 2.78 (SD = 0.83) after the program. Accordingly, the total score of depression increased from 33.00 (SD = 7.26) to 36.67 (SD = 6.95). As such, depression increased overall in the control group after the program, although not at a significant level.

On the other hand, the comparison of the pretest-posttest scores of the experimental group showed that the detailed items decreased in the pretest: depressive emotions from 13.50 (SD = 5.56) to 13.10 (SD = 4.89), positive emotions (reverse scoring) from 8.50 (SD = 2.42) to 7.90 (SD = 2.33), physical inactivity from 12.90 (SD = 2.92) to 11.80 (SD = 3.33), and interpersonal relationship from 3.60 (SD = 1.96) to 3.10 (SD = 1.91) after the program. Accordingly, the total score of depression decreased from 38.50 (SD = 11.756) to 35.90 (SD = 11.80). As such, the total score of depression and detailed items decreased overall in the experimental group after the program, although not at a statistically significant level.

Depression is a common psychological and emotional problem of caregivers caring for patients and suffering from caregiver burden, and this is especially more severe among caregivers of the elderly with dementia than other caregivers (Givens et al., 2014; Hernandez and Bigatti, 2010; Kuo et al., 2016). Depression suffered by dementia caregivers increase due to the behavioral and psychological symptoms and problems that come from the cognitive and behavior disorders of dementia patients (Givens et al., 2014). Accordingly, psychotherapeutic interventions for dementia caregivers are focused on social support as well as understanding the behavioral problems of dementia patients, learning how to manage them, and improving the individual competencies of caregivers (Kuo et al., 2016; Livingston et al., 2013; Yang et al., 2019).

Horticultural therapy is proved by many studies to be effective in improving depressive symptoms of various subjects. Gonzalez et al. (2010) implemented a horticultural therapy program on patients clinically showing depressive symptoms and discovered that nature-based horticultural activities bring fascination toward nature, helping them psychologically break free from the struggles in daily life, restore attention and improve clinical depressive symptoms. Gonzalez et al. (2011) also proved that the horticultural therapy program alleviates depression, anxiety, stress and negative emotions of adults diagnosed with depression and improves positive emotions, thereby reducing psychological pain. Moreover, according to literature reviews on horticultural therapy performed on subjects with mental and behavior disorders or problematic symptoms, horticultural therapy had positive effects on depression, which is one of the major symptoms of mental and behavior disorders (Clatworthy et al., 2013; Kamioka et al., 2014; Soga et al., 2017).

Unlike the results of previous studies in which horticultural therapy is found effective in alleviating depressive symptoms, this study did not show a significant level of positive change as a result of the statistical testing. This may be firstly due to the fact that the number of subjects is not big enough to test the effects of horticultural therapy. A non-parametric test was conducted to make up for this deficiency, but the problem is that statistical significance is dependent on sample size (Wasserstein and Lazar, 2016). The result of this study showed that, although not statistically significant, the control group showed an increase in scores overall, while the experimental group showed a decrease. This implies that different results may be expected when more subjects are recruited in further research. The second reason may be the fact that the program was carried out in a short period of time. The total scores of depression before the program showed that both the experimental group (M = 38.50, SD = 11.75) and control group (M = 33.00, SD = 7.26) showed high depression in the range of ‘Has depressive symptoms (22–60 points)’. According to Son et al. (2006), long intervention periods are effective if the purpose is mental and psychological rehabilitation, and Kim et al. (2020) also mentioned that it was effective to apply at least 3–4 months of the program aimed at mental and psychological rehabilitation. However, this study was conducted in a relatively short period of time, twice a week for 4 weeks (about a month), which is not enough time alleviate the high level of depression among the participants. However, the tendency of change in the depression scores between the groups showed that the depression scores increased in the control group whereas they decreased in the experimental group. This implies that different results may be expected when the program is implemented for a longer period of time in additional research.

As a result of implementing the horticultural therapy on dementia caregivers, the depression scores were decreasing overall, although not at a statistically significant level. Considering the effect of horticultural therapy in improving depressive symptoms as reported by many previous studies, the activities in the program enabled the participants to understand more about dementia, obtain social support by communicating with other dementia caregivers, and secure a positive self-image, which had effects on relieving their depressive emotions, establishing positive emotions, improving symptoms of slowing physical behaviors, and releasing the need for interpersonal relations, thereby alleviating depressive symptoms. Accordingly, this proved the applicability of the horticultural therapy program as an intervention for alleviating depressive symptoms of dementia caregivers. Further research is suggested to make up for the deficiency of sample size, which is the limitation of this study, and then test the additional issue of intervention period.

We tested the posttest mean difference of QOL between the groups after the horticultural therapy program and conducted the pretest-posttest mean analysis by group (Table 4). The result of testing the posttest mean difference between the groups showed that there was no statistically significant difference in the total score of QOL and all detailed items.

As for the pretest-posttest mean analysis by group in the control group, there was no statistically significant difference found overall. On the other hand, in the experimental group, general QOL and general health increased from 11.60 (SD = 3.75) to 14.20 (SD = 2.90) at a significant level (p = .020), and physical health increased from 13.09 (SD = 1.80) to 14.17 (SD = 2.31) at a nearly significant level (p = .057). Psychological health, social relations and environment also increased from 11.87 (SD = 2.84) to 13.80 (SD = 2.65), 12.53 (SD = 2.44) to 12.93 (SD = 1.55), and 12.50 (SD = 3.47) to 13.75 (SD = 2.90), respectively, but not at a statistically significant level. Moreover, as for the total score of QOL, the control group showed a decrease from 65.33 (SD = 9.23) to 64.16 (SD = 9.20), although not at a statistically significant level, whereas the experimental group showed an increase from 61.59 (SD = 12.98) to 68.85 (SD = 9.83) at a nearly significant level (p = .059).

Caregivers of the elderly with dementia face severe stress and constraints on daily life due to behavioral and psychological problems and difficulty in performing daily activities due to the cognitive impairment of the dementia patient, and their QOL decreases as a result (D’Onofrio et al., 2015; Kim, 2017). Therefore, Yang et al. (2019) claimed the need for occupational therapy intervention programs to maintain and improve the QOL for caregivers of the elderly with dementia.

Horticultural therapy is also one of occupational therapy interventions, and many studies are conducted to maintain and improve the QOL for various subjects. Oh et al. (2012) discovered that the horticultural therapy program for women that survived cancer was effective in reducing state anxiety and improving QOL. Sommerfeld et al. (2010) claimed that horticultural activities carried out as for recreation satisfied the desire for creation upon encountering nature and formed a healthy lifestyle, improved the QOL and health, and thus showed effect in increasing the elderly’s physical activities and QOL. Perrins-Margalis et al. (2000) applied horticultural therapy as a group activity for patients with chronic mental illness, and reported that it had an immediate positive effect on QOL such as life satisfaction, well-being and self-concept.

As such, many studies are proving that horticultural therapy is effective in improving QOL, and this study also showed that it had positive effects on promoting QOL for dementia caregivers. In terms of personal crisis intervention and interpersonal crisis intervention in each step of the program, the participants could interact with others and promote understanding about themselves in mutually dynamic connections with nature mediated by horticulture so that they can establish a positive self-image and rebuild interpersonal relations, thereby achieving self-growth and improving QOL. Furthermore, the process of horticultural activities increased physical activity and promoted physical health, and decorating the surroundings with outputs they had made themselves during the activities improved their living environment, thereby promoting QOL. In particular, the aforementioned effect of the intervention applied in this study is supported by Töpfer and Wilz (2020), who reported that the psychological intervention program for dementia caregivers improves positive experience based on the strengths of the subjects with cognitive-behavioral methods and teaches how to cope with negative experience, which serves as a potential mechanism for improving QOL.

We tested the posttest mean difference of caregiver burden between the groups after the horticultural therapy program and conducted the pretest-posttest mean analysis by group (Table 5). The result of testing the posttest mean difference between the groups showed that there was a statistically significant difference only in the score of social activity constraints, which is a detailed item of caregiver burden (p < .001).

As for the pretest-posttest mean analysis by group in the control group, there was no statistically significant difference found in the total score as well as detailed items other than social activity constraints. Social activity constraints showed a significant increase from 18.44 (SD = 6.21) to 23.11 (SD = 2.80; p = .050). This increased burden of social activity constraints some time after the pretest is explained by the fact that personal time of caregivers decreases in the process of caring for dementia patients, which reduces their time for leisure and social activities and limits social activities, thereby decreasing their satisfaction with interpersonal relations after some time (Chenoweth and Spencer, 1986; Deimling and Bass, 1986; H.J. Kim et al., 2016).

On the other hand, in the experimental group, the total score of caregiver burden decreased from 88.30 (SD = 13.47) to 82.50 (SD = 12.16), and in the detailed item pretest-posttest score change, social activity constraints decreased from 17.50 (SD = 3.17) to 13.60 (SD = 4.43) at a significant level (p = .028). This result is similar to that of many studies proving that many subject groups showed social effects of positive change in interpersonal relations, communication and sense of closeness through horticultural therapy (Harris and Trauth, 2020; Hartig et al., 2014; Kim et al., 2019; Song et al., 2015). Such social effects of horticultural therapy appear due to interactions in the dynamic process of horticultural therapy (Relf, 1981). Relf (1981) argued that interactions in terms of therapeutic dynamics cause various social contacts among subjects, between subjects and therapists, and between subjects and the community. In particular, the horticultural therapy program implemented in this study encouraged the participants to build rapport using the common topic of ‘dementia’ from the beginning of the program, which quickly promoted cohesiveness among the self-help groups and stimulated ‘interactions between participants’. Yet, due to the nature of the elderly with dementia, the burden of social activity constraints caused by the constraints on personal life may decrease only by participating in the existing family support program, breaking free from those constraints even for a little while (Seo et al., 2018). However, compared to other general results of this study, the statistically significant difference in the burden of social activity constraints also proved by the posttest analysis results between the groups may be the effect of horticultural therapy that provided interventions using horticultural activities as the medium with reference to applications of existing intervention programs for caregivers of the elderly with dementia. This is because horticultural activities help build smooth relationships with others in a nonverbal way even in terms of interactions among the subjects, thereby providing an ideal environment that encourages them to make social contact without resistance, such as facing the problem of interpersonal relations (Relf, 1981).

Moreover, although not statistically significant, the results also showed decreased scores in the negative change in the elderly-primary caregiver relationship, negative change in family relations, and psychological burden. This result is due to the gender characteristics, as the caregivers of the elderly with dementia are women. According to Iavarone et al. (2014), female caregivers suffer emotion-focused burden and pain as a coping strategy to protect the elderly with dementia, which can be reduced by providing restructuring of the cognitive view of dementia. Due to these characteristics, in the process of talking about their difficulties regarding dementia and obtaining knowledge about dementia in the horticultural therapy program, the participants built rapport and emotionally communed with others, and reestablished the meaning of family relations and relationship with the dementia patient, which restructured the cognitive view and reduced the scores in the emotion-focused burden. This positive change, although not statistically significant, may be due to the small sample size or short period of time previously mentioned as the limitations of this study. Accordingly, further research is suggested to make up for these deficiencies and, as mentioned by Iavarone et al. (2014), test whether the results are due to the difference in coping strategies to protect the elderly with dementia depending on gender as a demographic characteristic.

On the other hand, burden of financial and economic activities and burden of health tended to increase. This is because caregivers faced greater burden in terms of development and physical aspects due to gender characteristics (Iavarone et al., 2014), and also because it becomes more difficult for them to conduct economic activities due to the nature of the disease (J.Y. Kim et al., 2016; Kwon, 2002; Yoo and Kim, 2004). However, the horticultural therapy program in this study was formed to threat psychological and emotional problems of the participants, and thus could not alleviate burden of financial and economic activities. By overcoming the limitations of this study such as the relatively short period of time and limited space of the site where the program was implemented and carrying out various horticultural activities for at least a certain period of time in a place where big movements can be made such as gardening, it would be possible to solve the problems of physical burden through horticultural therapy, considering the results of studies by Sommerfeld et al. (2010) and Park et al. (2014) claiming that horticultural activities such as gardening prevented decline of physical functions due to aging and promoted physical health.

The intensified caregiver burden due to chronic stress of caregivers of the elderly with dementia has negative effects on psychological and physical health of not only the dementia patient but also other family members and caregivers, which is why there is a need for intervention strategies to support them in lessening their burden (Schulz et al., 2020). A meta-analysis proved that non-pharmacological intervention programs for caregivers of the elderly with dementia such as mindfulness-based counseling and psychotherapy were effective in reducing depressive symptoms, stress and caregiver burden of subjects (Cheng et al., 2020). Like the results of non-pharmacological intervention programs applied as psychotherapy, horticultural therapy also proved its effects in lessening caregiver burden, especially the burden of social activity constraints, of dementia caregivers based on the results of this study. Furthermore, positive results can be found in other categories of caregiver burden by applying a program that overcomes the aforementioned limitations of this study. The results proved the potential of horticultural therapy as an intervention strategy to lessen the caregiver burden of dementia caregivers based on the effects of alleviating the burden of caring for the elderly with dementia.