Stress is often defined as the psychological and/or physiological response to perceived threats or other internal/external stressors (APA, 2018a). It impacts emotions, influences behavior, and encompasses changes that affect almost every bodily system (Anisman and Merali, 1999; APA, 2018a). According to the Spring 2022 National College Health Assessment, 75% of students reported moderate or severe psychological stress. Stress was reported as an academic impediment by 37.3% of all students in 2023 compared to 32% in 2016, representing a 17% increase (ACHA, 2016, 2023). Not only can stress negatively impact academic performance (Córdova et al., 2023), but students experiencing a high level of stress also report reduced enjoyment in learning (Flinn et al., 2016), poorer skill acquisition (Córdova et al., 2023), and decreased memory retrieval (Córdova et al., 2023; Wolf, 2017).

The American APA (2018b) defines anxiety as an emotion characterized by worry, tension, intrusive thoughts, and physiological changes such as increased blood pressure and heart rate. Related to fear, anxiety tends to be correlated with a future-oriented focus on a broad or diffuse threat. Repeated patterns of anxiety can lead to sleep problems (Dunn et al., 2022), negative health outcomes (Aquin et al., 2017; WHO, 2023), and decreased cognitive performance (Majeed et al., 2023; Hartanto and Yang, 2022). In 2023, 30.4% of students reported having a diagnosed anxiety condition compared to 24.9% in 2016, representing a 22% increase (ACHA, 2016, 2023). Researchers found that high academic distress was associated with a higher probability of student anxiety (Zakeri et al., 2021). Anxiety is also positively associated with student burnout (Abreu Alves et al., 2022), reduced academic engagement (Mou et al., 2022), and lower academic performance (Moreira de Sousa et al., 2018).

Resilience encompasses three key components: an adversity, which constitutes a stressful or traumatic experience; observable signs of healthy functioning following the adversity; and mechanisms that prevent or facilitate recovery from the typical distress. These mechanisms, also referred to as the protective factors, enable the individual to recover from adversity (Hamby et al., 2017). The concept of resilience was originally conceptualized as individual, intrinsic traits such as coping skills and positivity. Further research on resilience increasingly acknowledges the protective role of extrinsic factors as well, including social relationships and resources within one’s environment (Luthar et al., 2000; Van Breda, 2001). Research based in social-ecological models has emphasized that resilience is characterized by complex interactions between the individual and specific characteristics of their environment. Researchers have proposed models that explore the qualities of different contexts that help facilitate resilience (Masten et al., 2023; Ungar, 2011; Ungar et al., 2013).

College is a time of significant transition to young adulthood (Walters et al., 2018), which can be challenging and stressful for many students as they experience changes in their environment, social and family relationships, education, responsibilities, and finances, among other areas. Resilience has been demonstrated to be an effective mechanism in empowering students to navigate these challenges, particularly in reducing stress and alleviating depressive symptoms (Smith and Yang, 2017). Despite these benefits, a meta-analysis by Chua et al. (2022) found that globally, 36% of college students reported low resilience. Students exhibiting low levels of resilience demonstrate poorer well-being, increased stress levels, a higher likelihood of dropping out of school (Ang et al., 2022; Sosu and Pheunpha, 2019), decreased academic motivation (Pascoe et al., 2019), impaired memory retrieval (Trammell and Clore, 2014), and poor academic performance (Chua et al., 2022). A study of first- and second-year medical students revealed that general distress and lack of adaptive coping strategies were associated with burnout (Abreu Alves et al., 2022). The study also reported that academic engagement served as a protective factor, reducing the effect of burnout on dropout intention.

Because education is a primary focus of the college experience, an area of resilience that is of particular interest is that of academic resilience. Academic resilience refers to the capacity to return to healthy academic or educational functioning following an adversity (Cassidy, 2016). Several risk factors may influence an individual’s academic performance including stress, anxiety, depression, and loneliness (Wohn and LaRose, 2014), as well as chronic illness, undesirable living conditions or homelessness, experiencing a natural disaster, and facing financial hardships among others (Horton, 2015; Masten, 1994; Sosu and Pheunpha, 2019; Yang and Wang, 2022). While numerous studies have been conducted on resilience and protective factors in mental health, research in protective factors influencing academic resilience remains limited.

Although the reasons for the dramatic increase in mental health challenges among college students during the past decade are not fully understood, contributing factors may include perceived increases in parental expectations (Curran and Hill, 2022), impacts of social media and social ‘worth’ (Powdthavee, 2014; Scelfo, 2015), financial stressors (McAlpine, 2021; NASPA, 2023), shifting societal values (McAlpine, 2021), and more recently, the trauma of Covid-19 (Kim et al., 2022; Lipson et al., 2022). On the more positive side, the awareness of mental health issues and reduction in the associated stigma may allow more students to voice their struggles (McAlpine, 2021; NASPA, 2023). Additionally, advances in treatment and medications have enabled more students with mental illness to go to college (Abrams, 2020; Biancolli, 2021).

College counseling and wellness centers everywhere are overwhelmed with almost 90% of center directors reporting an increase in the number of students seeking services in 2019 (LeViness et al., 2019). This high demand leads to longer wait times for students, mounting pressure on counseling centers, and counselor burnout (Abrams, 2020; Biancolli, 2021). However, colleges have unique access to students and therefore have greater opportunity to provide resources and additional interventions to those in need. As Lipson et al. (2022) point out, colleges and universities are ideal settings to approach mental health issues during a significant stage of life.

Recruitment for this study took place via social media posts, campus flyers, email announcements, and information provided at the UF Counseling & Wellness Center (CWC). Interested students were invited to email or call the study PI. During that initial contact the PI and/or study staff inquired about student status. If the individual was a current student at UF they were invited to join the group and informed of the research study. Students were not required to take part in the study to join the TH group but were invited to join the study if there were at least eight weeks remaining in the current semester.

Eighty-five participants from the student body of the University of Florida consented to participate in this research study (IRB# 202101905). The study took place in three phases over three semesters; Fall 2021, Spring 2022, and Fall 2022. Fifty-one participants who attended at least eight sessions and completed both the pre-survey and post-survey are included in the analyses here.

Students were admitted to the TH group on a rolling basis if there was space but invited to participate in the study only if there were at least eight sessions remaining in the semester. In addition, some students declined to participate in the research study so there was a mix of research and non-research participants in each group. All participants took part in the same activities no matter their status.

Sessions were offered at different times during the week (mornings and afternoons) to accommodate academic schedules. Fall cohorts began in September and ended in December. Spring cohorts began in January and ended in April. Each session lasted 75 min and followed the same schedule, which included 15 min of ongoing propagation work in a mini tabletop greenhouse, 20 min exploring a wellness strategy, and 40 min engaged in a horticulture activity. The structure and schedule of each session was the same throughout the groups. By the end of their eighth session, each participant had been introduced to eight different wellness strategies and participated in eight different horticulture activities. The underlying goals of each session were to: (1) build skills and confidence in working with plants; (2) introduce and practice a wellness strategy; (3) increase understanding of the role that plants and nature can play in personal wellness; and (4) build community among participants.

Each group met in the same greenhouse located on the University of Florida campus in Wilmot Botanical Gardens. The greenhouse is a 2,700 sf conservatory-style glass greenhouse with bathrooms, accessible features, and a fully air-conditioned reception area connected to the greenhouse. The greenhouse has a large area with tables and chairs where a group can work and interact comfortably. Most of the horticulture activities took place in the greenhouse while most of the wellness strategies were explored in the quieter reception area.

The greenhouse sits in the SW corner of the 4.8-acre gardens which contain a variety of garden features and plant types in both sunny and shady areas. The garden was used as an integral part of several horticulture activities such as collecting natural materials to use in projects. A few wellness strategies also took place in the gardens such as mindful walking.

Participants were assigned to a TH group with 6–12 other participants based on the group that worked best for their schedule. TH sessions were facilitated by a professionally registered horticultural therapist and a program manager with training in the use of horticulture as a therapeutic modality.

During each session, participants explored a wellness strategy, led by a counselor from the Counseling & Wellness Center at UF, and then engaged in a related horticulture activity. Wellness strategies included mindfulness, cognitive defusion, anxiety reduction techniques, and self-compassion strategies, among others. Horticulture activities included hands-on basic plant care such as plant propagation techniques and methods, proper planting and maintenance, and general plant identification and knowledge. Participants also participated in plant art activities and regularly took plants and plant crafts home.

This study used a single-group, non-experimental pre−/post-survey method. At the first visit, participants completed the informed consent. After signing the ICF, they filled out a demographic questionnaire inquiring about their age, race/ethnicity, gender, academic major, and gardening experience, among other information. Participants then filled out a set of four self-report psychometric questionnaires and then immediately joined the TH session with the rest of the group.

The therapeutic horticulture sessions ran for 12–13 weeks each semester. Study participants were asked to attend sessions once per week. If a participant was not able to attend at least eight sessions, their participation in the study was discontinued. The participant could continue attending the sessions, but any data collected was not used in final data analyses. At the participant’s eighth visit, following the scheduled TH activity, they completed the same four psychometric surveys as during their first visit. Participants who completed their eighth visit and corresponding questionnaires could continue to attend the remaining sessions, but no further data was collected from these individuals.

Stress. Participants’ perceived stress was measured using the Perceived Stress Scale (PSS) (Cohen et al., 1983), which is based on a 5-point Likert Scale and is the most widely used psychometric tool to gauge the perception of stress. This study used the validated 10-item Perceived Stress Scale, which has strong validity and reliability (Roberti et al., 2006).

All outcome measures were treated as continuous variables and paired sample t-tests were computed to determine if there was significant change on each outcome variable from the pre-survey to the post-survey. All analyses were conducted using the Statistical Package for the Social Sciences (SPSS), version 29.

Fifty-one individuals, who completed eight therapeutic sessions and also completed the pre-survey and post-survey, are included in this study. Study participants were currently registered students at the University of Florida, representing 34 different majors. The participants ranged in age from 18–53, with a median age of 22. This includes roughly equal numbers of undergraduate (52.9%) and graduate students (47.1%). About 71% of participants identified as female, 24% identified as male, and 6% identified as queer / gender nonconforming. Demographic and background characteristics are reported in Table 1.

This study has several limitations. First, there was no random assignment in this study. Participants signed up for the study knowing there were horticulture activities involved. Therefore the sample may be subject to selection bias since it is possible that some participants may have had a positive attitude toward the intervention. Second, the study did not include a control group to determine if the reported changes were in fact due to the TH intervention or other external influences during the semester. Third, all outcome measures used (ARS, PSS, BRS & STAI-6) rely on self-reported data, which could be influenced by social desirability bias, leading participants to over-report improvements or under-report negative emotions. Fourth, while race was not a focal variable of this analysis, the relatively high frequency of ‘No answer’ responses on the race question makes it more challenging to fully describe the characteristics of this sample. Lastly, the participants were all students enrolled at a single large public university and therefore the findings may not be generalizable to students at other institutions. Nevertheless, this research offers valuable insight into the potential benefits of therapeutic horticulture and serves as an important foundation for future studies on the impacts of TH on the mental health and academic resilience of college students.