Equine-assisted psychotherapy (EAP) has gained attention in recent years as a possible modality for promoting psychological, emotional, and social well-being among children and adolescents (1, 2). Theoretically, there is good reason to think that EAP might benefit struggling youth. Since equines are sensitive to nonverbal behavior and provide immediate feedback to handlers, learning to communicate with and guide a horse can facilitate increases in self-awareness and self-regulation (3, 4). Others posit that self-esteem and self-image are strengthened by grooming and caring for horses because participants adopt the role of responsible caretakers rather than patients when engaging in this type of therapy (5). EAP may be particularly helpful for youth with a history of trauma; While it can be challenging for these children to form healthy therapeutic relationships (6–8), engaging with animals provides a potential pathway for circumventing this barrier (3, 9, 10). Finally, working directly with equines may produce calming autonomic system responses; indeed, the release of oxytocin and inhibition of cortisol have been observed after equine interactions (11, 12).

Many studies indicate that equine-assisted programs are beneficial for youth at risk of poor outcomes, including those who have experienced abuse, exhibit behavioral problems, or meet the criteria for depression, anxiety, or PTSD (13–17). Improvements in self-efficacy and self-esteem (18, 19), and improvements in coping skills and emotion regulation have been observed following equine interactions in at-risk youth (20, 21). EAP may also improve social functioning and interpersonal relationships. For example, among adolescents with mental and behavioral concerns, weekly group equine-assisted learning sessions contributed to significant reductions in anti-social behavior and improved socialization skills (22). Clinically meaningful reductions in depression and anxiety have also been observed following EAP in this population (23–25).

However, very few EAP protocols have included a comparison group, limiting our ability to understand the value of EAP relative to traditional methods for treating psychosocial concerns in at-risk youth. The majority of recently published findings for equine interventions used to improve youth emotional well-being did not include a control or comparison group of any kind (10, 17, 18, 22, 23, 26–34). Those that do offer a comparison group usually compare equine-assisted therapy with a group of waitlisted control participants (13, 16, 20, 25, 35–37). The lack of controlled studies limits our understanding of the utility of equine-assisted therapies relative to standard psychotherapy.

Only a small set of recent studies directly compare the psychosocial benefits of EAP with standard treatment for improving emotional well-being in samples of youth. Authors of a recent meta-analysis of equine-assisted interventions for youth found just 13 studies published in the last 25 years that included credible comparison treatments (2). While their results revealed significant improvements in externalizing, and internalizing behaviors and efficacy (but not depression or self-esteem) following equine-assisted therapies, the team cited the need for additional reports with larger samples, randomization, or at least tests for equivalence between groups at baseline, and comparison groups with credible treatments instead of waitlisted controls.

While EAP is increasingly understood to be an effective therapeutic technique, it is less clear how the efficacy of such therapy compares to standard outpatient psychotherapy for youth at general risk of poor outcomes. Further research on its effectiveness is needed to determine if investments in and referrals for EAP are well placed. Thus, we evaluated the efficacy of EAP compared to TP for improving psychosocial outcomes in youth referred for psychotherapy.

Global psychological, social and emotional well-being scores improved from baseline to follow-up by 18-23% among all participants (see paired-t test results in Table 2 ). The adaptive perspectives and cognitive skills measured also improved significantly (12% to 28%) with one exception; average emotion suppression scores did not change across the treatment period (p = .77).

As expected, adolescent clients experienced significant psychosocial benefits from receiving community-based non-profit therapeutic services. On average, clients experienced 18-23% gains in self-reported ratings of psychological, social, and emotional well-being after engaging in services ( Figure 1 ). Notably, clients self-reported that their hope and resilience scores, strong predictors of success-oriented behaviors and positive adjustment in youth (49, 50), improved following services. A 22% increase in cognitive reappraisal skills, known to be associated with enhanced well-being and prosocial behavior (51–53), was also observed following treatment. These data are consistent with a large body of work that demonstrates that psychotherapy is effective at improving the well-being of adolescent clients (54) and that community-based psychotherapy programming is effective (55).

Improvements in all psychosocial outcomes were similar for those engaged in EAP and TP. Psychosocial improvements following EAP might be the result of confidence and emotion management developed in the horse arena, crossing over into other areas of life (33, 56–58). While these data cannot corroborate or disprove mechanistic theories, they do indicate that using EAP as the primary treatment mode provides similar benefits for struggling youth as traditional psychotherapy. Notably, in this nonrandom design, a history of nonresponsiveness to past TP was one criterion for placing clients in the EAP group. Despite this bias, rates of improvement during treatment were similar between the groups. Likewise, the EAP group was slightly younger and had slightly better cognitive reappraisal scores than the TP group at baseline. Despite these baseline differences, rates of improvement were similar between the two groups.

Notably, clients in neither group experienced desired reductions in emotion suppression. Neuroimaging studies show that emotion suppression is cognitively taxing and activates stress physiology that can negatively influence immune function and physical health (59). Since high levels of emotion suppression create significant psychological distress (60), emotion regulation strategies that reduce suppression serve to promote resilience through difficult childhood experiences (61). Congruently, replacing maladaptive emotion regulation strategies with adaptive ones is encouraged as a central goal of youth treatment programs that seek to promote resilience (62). Given that we observed improvements in outcomes that are known to covary with emotion suppression (resilience, cognitive reappraisal), and that others have reported suppression to be modifiable in youth (63) the lack of change in this area is difficult to interpret.

Notably, most research connecting suppression to negative outcomes refers to the suppression of negative emotions. The scale used herein included questions related to suppressing both negative and positive emotions, the latter of which may not be problematic (64). In fact, from an adolescent perspective, the ability to suppress positive emotions (ie., laughter during a school session) may be beneficial and related to improved social awareness, engagement, and desirable emotion regulation (65). Indeed, recent factor analyses of an expanded version of the scale provided estimates that support separating the measurement of suppression for positively and negatively valenced emotions and indicate that they differentially relate to mental health (66). Thus the measure utilized in our study, while psychometrically sound, may not be an optimal indicator of psychosocial health. Additionally, any reductions in negative emotion suppression may have been counterbalanced by potential improvements in adaptive positive suppression.

Generalization of these results are limited by varied levels of client engagement and the use of non-random assignment. This analysis was limited to clients who completed both intake and follow-up surveys within the study period. Clients who failed to complete data collection instruments may have differed from those who did. For example, extreme distress at the first visit may reduce the likelihood of clients completing intake data. The number of clients that the organization served but failed to complete the surveys necessary to be represented in this sample is unknown. Therefore, this data does not represent how the average client in the program fared; rather, it represents how clients who completed both survey instruments responded to treatment. Second, instead of using random assignment, the clinical manager made treatment decisions based on need and fit, an approach that, while methodologically limiting, was ethically aligned with this organization’s mission.

As expected with non-random procedures, the groups differed systematically in some ways. The EAP group was younger and had better cognitive reappraisal skills at program entry than the TP group. They may have also differed in other, unassessed ways. For example, given that EAP and TP are offered at different sites, the proximity of the client’s residence to a specific site could have influenced retention (and therefore inclusion in this analysis). Clients with greater parental support or high socioeconomic status may have been likely to successfully adopt a treatment plan that required regularly traveling to the more rural EAP center (note that if staff were aware of cost-related transportation barriers, gas cards were provided). Neither the client diagnosis nor the concurrent use of complementary treatments were documented for this study. These characteristics may have differed by group. Despite the few known and possible unknown baseline differences, the rate of psychosocial improvement in the two groups was the same. That is, clients in each group improved at the same rate regardless of their starting point. This ecologically valid method of treatment assignment resulted in similar improvements across both groups. However, generalizations should be made cautiously since non-random assignment may be associated with inflated effect sizes (67).

Additional limitations include the self-reported nature of the data and the lack of data related to dropout rates, unknown diagnoses and cultural diversity. Subjective differences between clients in perceived improvement and well-being may differ greatly, and the collected data may suggest perceptual improvements rather than empirical improvements to psychosocial well-being. Future studies may seek to incorporate data from clinician assessments, although this method of assessment is costly and time-consuming. To account for this obstacle, data collection from other informants such as parents or teachers may provide sufficient information. Qualitative and mixed-method designs should be employed in future studies to mitigate the subjectivity of client perception (23). Specific diagnoses or reasons for seeking treatment were not available to the research team. As a result, highly desirable subgroup analyses to determine potential therapeutic targets were not possible. Given the demographics of the communities this organization serves, the sample was likely predominantly Caucasian. We did not have the data necessary to assess differences in outcomes based on race or ethnicity, though such work is important (14).

These data indicate that community-based non-profit programs that offer therapy delivered by licensed therapists benefit at-risk youth. We provide evidence that youth with psychosocial challenges may receive similar benefits from treatment plans that primarily focus on consistently delivered, individualized EAP sessions as clients whose treatment plans are structured around individualized TP.