Dog-Assisted Interventions Reduce Salivary Cortisol in Ukrainian Military Personnel with Post-Traumatic Stress Disorder (PTSD): A Pilot Study

Abstract

Although recent studies report elevated rates of post-traumatic stress disorder (PTSD) among Ukrainian refugees, data from military hospitals and rehabilitation centers within Ukraine remain scarce. Numerous studies have described that interactions with therapy dogs help mitigate PTSD symptoms by providing emotional support, reducing hypervigilance, encouraging social engagement, and facilitating physiological de-arousal. However, the potential role of dog-assisted interventions (DAIs) in therapeutic settings during ongoing conflict has not yet been investigated. In this study, saliva samples pre- and post- a 20 min DAI were gathered from soldiers and other military service personnel with diagnosed PTSD in Kyiv, Ukraine. Salivary cortisol was assessed as a biomarker to parallel acute physiological strain. The findings revealed a decrease in salivary cortisol from pre- to post-DAIs in both men and women. Of note, men exhibited higher salivary cortisol levels than women both pre- and post-DAIs. The present findings suggest that people with PTSD benefit from canine support due to reduced acute arousal in a war environment. Limitations include the small sample size and the lack of a control group. In order to mitigate against the high prevalence of psychophysiological distress in at-risk populations such as military personnel, further research is warranted.

1. Introduction

War and crisis are significant public health concerns with immense and unavoidable short-term and long-term mental health implications. The World Health Organization (WHO) estimates the frequency of mental disorders in a war-exposed population to stand at 22.1% [1], with anxiety, depression, and post-traumatic stress disorder (PTSD) being the most prevalent conditions reported [2,3]. The majority of previous studies focused on examining the psycho-emotional state and mental health in post-war periods [4,5,6]. These reports demonstrated that exposure to traumatic events, particularly those involving threats to one’s own life or bodily integrity, or witnessing harm to others, significantly increases the risk of developing mental health disorders [7,8,9]. Individuals who experience war, directly or indirectly, are vulnerable to an array of psychological and physical health consequences [10,11]. Hoppen & Morina (2019) reported that 24% of adult war survivors met diagnostic criteria for PTSD and 23% for major depression, and of these, approximately 50% met the criteria for comorbid PTSD and depression [11]. These findings concur with others who reported that approximately 30% of a population exposed to war and displacement suffer from PTSD and/or major depression [9]. Given the high incidence and debilitating nature of PTSD, it warrants particular attention.

The global International Classification of Diseases (ICD) 11 model for PTSD includes six symptoms consisting of three core elements, composed of two symptoms: (a) re-experiencing the trauma in the present, (b) avoidance of traumatic memories, and (c) a persistent sense of threat that is manifested by increased arousal and hypervigilance [12]. Frequent reactions to traumatic stimuli include fighting or fleeing, or freezing, which represent reactions rooted in the body’s natural survival mechanisms. However, the specific nature of an individual’s response may vary widely depending on several factors, including age, prior exposure to trauma, personal characteristics, socioeconomic background, and genetic predisposition to stress-related disorders [13,14]. Population-based data suggest that approximately 36% of Ukrainian adults meet at least one criterion for a mental health disorder [14].

Notably, civilians affected by war frequently face greater psychological susceptibility than military personnel, who may be better prepared for combat-related stress [15,16]. Recent studies indicate that Ukrainian refugees—even though not engaging directly in warfare—report higher levels of psychological distress than military combatants [17,18,19], underscoring the profound and widespread mental health consequences of war. An extensive number of Ukrainians, who have taken up arms and joined the armed forces, are civilians, however [20,21]. On the one hand, actively defending one’s country has the power to confer a profound sense of purpose and meaning, potentially exerting a protective effect on psychological welfare. On the other hand, civilian combatants may face an elevated risk for injury because of the rapid deployment into combat roles with minimal or no prior training. Both insufficient preparation and lack of operational experience may heighten their vulnerability to stress reactions [20,22,23]. Moreover, the integration of civilians into military service frequently occurs without systematic screening for psychological or physical suitability, thereby amplifying the peril for mental health problems [20,24,25]. Civilian combatants, like professional military personnel, are likely to experience separation from their families and social networks—a factor known to substantially increase the likelihood of mental health difficulties, given that social connectedness constitutes a potent buffer against stress [26,27]. Women, older adults, and those displaced or living in occupied territories reported particularly elevated symptoms of PTSD [21].

Stress is a normal and adaptive psychophysiological response to actual and perceived threats, thereby assisting an organism to deal with challenges [28]. It is transient and resolves once the stressor is removed. Stress exerts its negative effects when it becomes chronic, overly intense, or poorly managed, leading to a profound dysregulation of the hypothalamic–pituitary–adrenal (HPA) axis, which starts with prolonged elevations followed by exhaustion of glucocorticoid hormones [28]. Cortisol plays a critical role in enabling individuals to adapt to stress, regulate metabolism, and modulate immune responses. However, its regulation is influenced by intrinsic physiological processes, social dynamics, and environmental conditions, making it a sensitive indicator of cumulative stress exposure [28]. Elevated stress levels have been documented as one of the key psychological impacts of the ongoing Russian invasion of Ukraine [29]. Acute stress is characterized by cascading emotional and physiological arousal occurring within the first month following exposure to a traumatic event. However, in some individuals, distress persists well beyond the initial incident [30]. Prolonged or repeated exposure to stressors such as violence, displacement, or bereavement can substantially increase the risk of severely impaired mental health over time, thereby priming PTSD [31,32]. Research comparing prisoners of war, active soldiers, and civilians highlights the coexistence of psychological distress and PTSD, illustrating individual differences in adaptation to repeated trauma [2,9,11,21]. These patterns of vulnerability are consistent with the polyvictimization model, which posits that exposure to multiple forms of victimization—such as witnessing violence, displacement, loss of loved ones, and personal harm—produces cumulative effects that increase susceptibility to PTSD, depression, and other stress-related disorders [11,25,31]. Collectively, these findings underscore the extensive mental health risks faced by military personnel and civilians within the context of ongoing armed conflict.

Kateryna et al. (2023) [33] reported that companion animals contribute significantly to the preservation of Ukrainians’ emotional and mental well-being, and Elmously (2025) [34] validated that pets hold a particularly important role in supporting the psychological resilience of displaced Ukrainians. Soldiers have a long and deeply rooted history of human–dog relationships [34], and dogs have had an array of roles, including scouts and couriers, but have also provided a source of comfort and emotional sustenance [35,36,37].

Dogs are widely recognized as beneficial to human health, specifically for individuals with PTSD, with documented positive effects on mental health, physical well-being, and overall quality of life [38,39,40,41,42]. Interactions with assistance or therapy dogs have been shown to help mitigate PTSD symptoms by providing emotional support, reducing hypervigilance, encouraging social engagement, and promoting physiological calming responses such as reduced cortisol levels and lower heart rate [42,43,44,45]. Behavioral activation through human–animal interaction is beneficial in the treatment of PTSD [46] and is an advantageous component of DAI. The presence of a dog may also contribute to the reduction in hyperarousal symptoms by helping ground the individual in the present moment, reinforcing a sense of safety. Dogs assist in diverting attention from negative thoughts, emotions, or intrusive memories. Consequently, they are regarded as calming agents that help regulate mental and emotional states in stressful situations [46,47].

Thus, the positive impact of DAIs is recognized, but preliminary research has failed to demonstrate that these positive effects are also attainable in exceptional situations like an ongoing war scenario. This study sought to examine the effects of a DAI on short-term salivary cortisol secretion in hospitalized military personnel in Ukraine.

2. Materials and Methods

2.1. Study Participants

This study included military personnel (n = 46, 33 men and 13 women) receiving treatment at a military hospital in Ukraine. All participants held a clinical diagnosis of PTSD according to standardized medical diagnostic criteria and confirmed by qualified mental health professionals. Age was calculated based on the written birthday on the documents at the time of the intervention. All records were screened for plausibility and completeness. One session’s data entry was marked as inconsistent with the study timeline and excluded. Gender was recorded as reported on the documentation. No randomization or control group was implemented due to ethical and operational constraints in a warzone clinical setting.

Recipients were treated in a military hospital in Kyiv at the time of recruitment and intervention. All participants were recruited by convenience sampling methodology. No further details are given to specifically protect the integrity of all participants of the intervention with regard to the ongoing war in Ukraine. Informed consent was signed by each study participant. Respective draft in both languages are available in the Supplementary Materials. Study participation was voluntary, and each participant could withdraw at any time without giving a reason.

2.2. Dog-Assisted Interventions (DAIs) and Saliva Sampling

DAI sessions were conducted between the 13th and 16th of January 2025. All participants attended a single recipient intervention session scheduled between 11 am and 2 pm to lessen potential confounding effects of circadian rhythmicity, similar to earlier experimental studies [48,49,50]. Session dates were recorded for each participant. DAI teams were recruited by a post via social media. Inclusion criteria for Ukrainian dog handlers were being at adult age (18+ years) and holding a valid DAI certificate, either awarded by an International Society for Animal Assisted Therapy (ISAAT) or European Society for Animal Assisted Therapy (ESAAT) [51,52]-recognized organization, or currently attending an ISAAT- or ESAAT-certified DAI education program. Together with their licensed dog handlers, 17 dogs in total participated in the DAI sessions of this study. Data on therapy dogs and their handlers were not a focus of this study and are therefore reported elsewhere [53]. The DAI sessions were structured animal-assisted activities for inpatients that lasted for 20 min each and included an introductory phase in which the recipient and therapy dog had the chance to approach each other, followed by an interactive sequence, in which calming or activating activities were carried out. The sessions were finished in the final sequence, where the dog handler provided positive feedback to the recipient and thanked them for compliance [53].

Saliva sampling was executed immediately pre- and post-session according to manufacturer recommendations (LABOklin saliva swabs, LABOklin, Bad Kissingen, Germany). To minimize contamination, food or drinks were withheld from each participant at least 30 min before saliva collection. Participants were asked to avoid caffeine prior to DAIs. To absorb saliva, the swab was gently inserted into the cheek pouch and held in place for approximately 40 s to soak sufficient saliva before being transferred to a collection tube, frozen, and stored until analysis. To determine salivary cortisol, an Enzyme-linked immunosorbent assay (ELISA) was used (Synevo laboratory (https://www.synevo.ua/, (accessed on 12 April 2025), Kyiv, Ukraine)). Cortisol concentrations (µg/dL) were measured according to the manufacturer’s standardized protocol.

2.3. Study Design and Statistics

A within-subject pre–post observational design was employed to evaluate changes in salivary cortisol linked to DAIs. For all demographic and session variables, descriptive statistics were calculated. Continuous variables are presented as means and standard deviations. Categorical variables are presented as frequencies and percentages. Data were outlined as mean ± standard deviation (SD) for normally distributed variables and as median with interquartile range (IQR) for skewed variables. Normality of cortisol data was assessed separately for females and males using the Shapiro–Wilk test as well as visual inspection of boxplots and histograms. Female cortisol values were approximately normally distributed (pre: W = 0.943, p = 0.326; post: W = 0.940, p = 0.301), whereas male cortisol values were non-normally distributed (pre: W = 0.873, p = 0.002; post: W = 0.903, p = 0.008). Within-group comparisons for females were analyzed through paired t-tests. For males, pre- versus post-intervention variations were analyzed using the Wilcoxon signed-rank test. Differences between females and males were evaluated using the Mann–Whitney U test for non-parametric data. An optional two-way mixed ANOVA was performed with Time (pre- vs. post-) as the within-subject factor and Gender (male vs. female) as the between-subject factor.

Statistical significance was set at p < 0.05. Cohen’s d was calculated for parametric tests; r was calculated for non-parametric tests. To analyze the effect of the intervention for the recipient group, we used a Paired Samples t-test to compare the levels of cortisol pre- and post-intervention. Cohen’s d (for t-test) or partial eta-squared (η_P² for ANOVAs) were calculated for the estimation of effect sizes. Established conventions for interpreting effect sizes are a Cohen’s d of 0.20 as a small, 0.50 as a medium, and 0.80 as a large [54,55], and η_P² values of 0.01 as small, 0.06 as medium, and 0.14 as large [55,56]. To assess the relationship between age and cortisol levels pre- and post-intervention, Spearman correlations were performed. All analyses were conducted using SPSS v28 (IBM, Armonk, NY, USA).

3. Results

3.1. Demographics

Study participants ranged in age from 18 to 63 years and were predominantly male (see

Table 1. Sociodemographic characteristics of the study participants.

Total Sample (n = 46)
Gender	n (%)
Male	33 (72)
Female	13 (28)
Sample	Age 1 Md (IQR)
All	35.6 (28.9–44.0)
Male	37.2 (33.8–45.0)
Female	29.0 (26.9–37.7)

¹ Participant age in years (at the time of the intervention session in January 2025) given in median (Md) and interquartile range (IQR).

). The mean age (± SD) of participants was 36.1 ± 10.2 years. The Shapiro–Wilk test was not significant (W = 0.97, p = 0.34), suggesting that the assumption of normality for the age of the recipients is given.

3.2. Within-Group Comparison

The descriptive data indicate that females exhibited lower cortisol levels than males at both sampling timepoints (see

Table 2. Descriptive statistics of salivary cortisol (in µg/dL) in males and females before and after dog-assisted interventions (DAIs).

	Mn ± SD	Median (IQR)	Min–Max
Male participants pre DAI 1	0.262 ± 0.172	0.225 (0.135–0.349)	0.067–0.470
Male participants post DAI	0.179 ± 0.145	0.143 (0.114–0.230)	0.022–0.831
Female participants pre DAI	0.131 ± 0.110	0.093 (0.086–0.164)	0.024–0.376
Female participants post DAI	0.081 ± 0.066	0.075 (0.042–0.121)	0.018–0.240

¹ Dog-assisted intervention.

). Statistical analyses indicate that salivary cortisol significantly decreased post-intervention in male participants (Wilcoxon signed-rank W = 570, p < 0.001, r = 0.61), indicating a large effect. Similarly, salivary cortisol also significantly decreased for female participants post-intervention, demonstrating a large effect (paired t-test t(12) = 4.12, p = 0.001, Cohen’s d = 1.14) (see Figure 1).

3.3. Between-Group Comparison

As depicted in

Table 3. Salivary cortisol concentrations (Mn ± SD) in males and females before and after dog-assisted interventions (DAIs) (p < 0.05 = *).

	Salivary Cortisol (µg/dL)	Test	p	Effect Size
Male vs. Female (Pre-DAI 1)	0.262 vs. 0.131	Mann–Whitney-U	0.031 *	r = 0.34
Male vs. Female (Post-DAI)	0.179 vs. 0.081	Mann–Whitney-U	0.042 *	r = 0.33

¹ Dog-assisted intervention.

, males displayed significantly higher pre-intervention baseline cortisol levels compared to the female group (Mann–Whitney U = 179, p = 0.031, r = 0.34). Also, male cortisol concentrations were similarly higher post-intervention compared to the female cohort (Mann–Whitney-U = 154, p = 0.042, r = 0.33).

The two-way mixed ANOVA showed a significant main effect of Time, confirming an overall salivary cortisol reduction (p < 0.001) and a main effect of Gender, stating higher cortisol levels in males (p < 0.05). However, the Time × Gender interaction was not significant, indicating similar decreases across Gender. Regarding age correlation, Spearman correlation analysis showed no significant effect between age and cortisol concentrations pre-intervention ρ ≈ −0.007, p = 0.963, and post-intervention ρ ≈ 0.013, p = 0.931.

4. Discussion

The present study investigated the effects of a single 20 min DAI session on salivary cortisol concentrations in hospitalized male and female Ukrainian military personnel. Exposure to armed conflict places both military personnel and civilians at substantial risk for psychological disorders [21,57]. Many Ukrainians have taken up arms, either formally joining the military or acting as volunteer civilian combatants. As a complementary mental health service, DAIs have emerged to mitigate stress and improve psychological well-being in trauma-exposed populations.

Prima facie, our data reveal that DAIs reduced cortisol levels in both male and female participants with large effect sizes, demonstrating that brief, structured interactions with therapy dogs attenuate acute stress responses in a high-risk population. Most existing research on the psychological impact of war has been conducted primarily in post-war contexts, so our data gathered in an active warzone address this gap in the literature.

Men exhibited higher baseline salivary cortisol concentrations and experienced a greater absolute decrease post-intervention. Women also showed a reduction in cortisol, although baseline levels were lower, which is consistent with prior literature suggesting differential stress physiology across gender [58]. Of note, previous research highlights that ongoing war exposure results in a down-regulation of long-term cortisol levels as determined in hair samples [59].

Despite these differences, the reduction in cortisol post-DAI was similar across genders. This is particularly interesting as studies indicate that women are generally more vulnerable to trauma with consistently higher levels of anxiety, depression, disturbances in self-organization (DSO), PTSD, and complex PTSD [7,35,60]. In Ukraine, women, older adults, and those displaced or living in occupied territories reported particularly elevated symptoms of PTSD [21]. Age was not associated with baseline or post-intervention cortisol levels, suggesting that the stress-mitigating effects of DAI are broadly applicable across adult age groups in this sample. These results suggest that DAI can effectively modulate physiological stress in male and female participants, even in populations exposed to high-stress environments. In line with the stress–diathesis model, mental health dynamically results from the interaction of environmental stressors and individual vulnerabilities [58,61].

Empirical studies support the role of dogs in promoting psychological resilience. Dogs provide support, serving as emotional anchor and practical aids, and their presence has been shown to alleviate stress and anxiety, fostering resilience and facilitating recovery [62,63]. Research indicates that the interaction with dogs may lead to a cortisol decrease, while simultaneously promoting the release of oxytocin, enhancing feelings of connection and safety [63,64]. Referring to the ongoing war, Kateryna et al. (2023) [33] reported that pets significantly contribute to the preservation of Ukrainians’ emotional and mental well-being. Also, Elmously (2025) validated animals as important in supporting the psychological resilience of displaced individuals [34]. Soldiers have a long and deeply rooted history of human–dog relationships, with dogs fulfilling operational roles (e.g., scouts, couriers, combat participants) as well as providing emotional support [40,41,65]. Mechanistically, dog interactions may reduce hypervigilance, promote social engagement, and trigger physiological calming responses, including reductions in heart rate and cortisol levels, consistent with the cortisol reductions observed in our study [42,43,45,47]. Behavioral activation, grounding effects, and distraction from intrusive thoughts further support dogs’ role in regulating emotional states, particularly in stressful and trauma-exposed environments [49,65,66,67]. Even brief animal-assisted interventions may produce measurable physiological benefits [67]; as suggested in this study, 20 min DAI sessions led to changes in acute arousal. Similarly, Haluza et al. (2025) [50] reported comparable reductions in salivary cortisol levels during a 20 min forest visit where individuals were quietly seated and observed the environment, compared to a control group who stayed in an urban area.

Our results further complement recent systematic reviews demonstrating the psychological benefits of DAIs in military and trauma-exposed populations. Chirico et al. (2022) [39] found that DAIs reduce PTSD symptom severity among military personnel and veterans, while Leighton et al. (2022) [41] reported that psychiatric assistance dogs yield clinically meaningful reductions in PTSD symptoms and improvements in mental and social health. Additionally, studies in Ukrainian populations during conflict indicate that companion animals contribute to emotional resilience and mental well-being, particularly in displaced civilians and high-stress military contexts [37,38]. Collectively, these data provide strong evidence that DAI may support psychological and physiological stress regulation in trauma-exposed populations. Thus, DAIs may be one vital resource to alleviate the psychological impact of the ongoing Ukrainian war for all populations involved. In conflict zones, DAIs may offer comfort and companionship, enhancing mental support, serving as one effective measure to address PTSD, anxiety, and depression.

The implications of our results are significant for both men and women in military and civilian mental health interventions. Acute stress regulation via DAIs may complement established psychotherapeutic approaches, offering a low-stigma, non-pharmacological adjunct that can be deployed even in operationally constrained environments. The large effect sizes observed suggest that DAIs could be integrated into broader military or civilian stress-management programs, particularly for populations at risk of PTSD or elevated stress due to combat, displacement, or other trauma exposure [67,68]. Of note, it would have been valuable to implement a control group that participated in alternative activities without a dog being present to demonstrate that decreasing cortisol levels were causally related to the DAI.

The generalizability of the present data is limited due to the pilot nature of the study and the constraints caused by the active conflict. Limitations of the study refer to the convenience sampling methodology, the use of salivary cortisol as the sole biomarker to parallel physiological arousal, and the relatively low sample size. Ideally, several follow-up sessions for each patient would have been desirable to investigate whether the observed patterns of lower salivary cortisol post-session are reproducible. It would have been interesting to complement the cortisol data with standardized scales assessing subjectively perceived stress and affective states. Also, qualitative data, such as visual analogue scales, would be desirable to complement the physiological data. Accordingly, future studies should consider additional biomarkers or other non-invasive procedures to monitor stress responses [69].

5. Conclusions

In conclusion, reduced salivary cortisol concentrations in both male and female participants provide preliminary evidence that brief behavioral interventions with therapy dogs can attenuate physiological stress in individuals diagnosed with PTSD. Our results support the potential role of DAIs as an effective non-pharmacological, low-stigmatized intervention for reducing physiological stress and supporting adaptation to adverse experiences in populations in high-risk environments, such as an active warzone. Our pilot findings clearly warrant further investigation. Future studies should include a bigger sample size and a suitable control group to obtain more robust results.

Existing research has not yet yielded conclusive evidence regarding the explanatory pathways or the specific physiological mechanisms through which benefits of human–animal interactions are conferred. A more comprehensive synthesis of the available evidence is therefore essential to understand the potential impacts of human–dog interactions on both physical and psychological health, as well as to identify the underlying mechanisms that mediate these effects. DAI programs could be modeled to be included in a structured, integrated health service, and stakeholders in Ukraine may provide much-needed additional support to those affected by the war, ultimately promoting their healing and recovery.