Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Feb 1.
Published in final edited form as: J Consult Clin Psychol. 2018 Feb;86(2):179–188. doi: 10.1037/ccp0000267

Preliminary efficacy of service dogs as a complementary treatment for posttraumatic stress disorder in military members and veterans

Marguerite E O’Haire 1, Kerri E Rodriguez 1
PMCID: PMC5788288  NIHMSID: NIHMS914972  PMID: 29369663

Abstract

Objective

Psychiatric service dogs are an emerging complementary treatment for military members and veterans with posttraumatic stress disorder (PTSD). Yet despite anecdotal accounts of their value, there is a lack of empirical research on their efficacy. The current proof-of-concept study assessed the effects of this practice.

Method

A non-randomized efficacy trial was conducted with 141 post-9/11 military members and veterans with PTSD to compare usual care alone (n = 66) versus usual care plus a trained service dog (n = 75). The primary outcome was longitudinal change on the PTSD Checklist, including data points from a cross-sectional assessment and a longitudinal record review. Secondary outcomes included cross-sectional differences in depression, quality of life, and social and work functioning.

Results

Mixed model analyses revealed clinically significant reductions in PTSD symptoms from baseline following the receipt of a service dog, but not while receiving usual care alone. Though clinically meaningful, average reductions were not below the diagnostic cutoff on the PTSD Checklist. Regression analyses revealed significant differences with medium to large effect sizes among those with service dogs compared to those on the waitlist, including lower depression, higher quality of life, and higher social functioning. There were no differences in employment status but there was lower absenteeism due to health among those who were employed.

Conclusions

The addition of trained service dogs to usual care may confer clinically meaningful improvements in PTSD symptomology for military members and veterans with PTSD, though does not appear to be associated with a loss of diagnosis.

Keywords: posttraumatic stress disorder, animal-assisted intervention, military veterans, service dogs, human-animal interaction

Posttraumatic stress disorder (PTSD) is a trauma and stressor-related disorder that adversely affects the mental health and quality of life of a substantial number of United States military members and veterans (American Psychiatric Association, 2013; Kang, Natelson, Mahan, Lee, & Murphy, 2003; Kulka et al., 1990). The traumatic event of experiencing combat violence associated with military deployment is particularly associated with a risk of developing PTSD, enough so that the recognition of the disorder by modern psychiatry in 1980 was largely brought about as result of the mental health experiences of military members returning from the Korean and Vietnam Wars (Trimble, 1985). An estimated 6–14% of all Operation Iraqi Freedom (OIF) and Operation Enduring Freedom (OEF) military veterans returning from deployments to Iraq or Afghanistan are affected by PTSD (Hoge et al., 2004; Tanielian & Jaycox, 2008). Further, PTSD in this population is linked to negative comorbidities such as depression (Grieger et al., 2006), alcohol and substance abuse (Jacobson et al., 2008), and suicidal behavior (Kemp & Bossarte, 2013).

The demand remains high for effective PTSD treatment options for military personnel. Current evidence-based treatments for PTSD are effective for many individuals; however, dropout and nonresponse rates can be up to 50% (Hoge et al., 2014; Mott et al., 2014; Schottenbauer, Glass, Arnkoff, Tendick, & Gray, 2008). These high dropout and nonresponse rates may be explained by barriers to receiving mental health care specific to the military population. Such barriers can range from conflicts with work, school, or family commitments to social stigmas and stereotypes surrounding treatment (Ouimette et al., 2011; Pietrzak, Johnson, Goldstein, Malley, & Southwick, 2015). Further, common symptoms of PTSD and depression such as denial, avoidance, and helplessness can exacerbate the problem (Sayer et al., 2009). It is therefore imperative to discover and evaluate alternative and complementary therapies (Bomyea & Lang, 2012; Cukor, Spitalnick, Difede, Rizzo, & Rothbaum, 2009). In particular, there is a need to evaluate complementary treatment options for PTSD that encourage engagement and retention while directly or indirectly addressing the comorbidities of the diagnosis (Hoge et al., 2014; Ouimette et al., 2011; Pietrzak et al., 2015).

One of these emerging treatment options is the placement of a specially trained PTSD service dog. Psychiatric service dogs are distinguished from emotional support, therapy, or companion dogs by specifically being trained to perform a variety of commands relevant to the psychiatric needs of the individual and thus are legally allowed public access under the Americans with Disabilities Act (ADA; Kruger & Serpell, 2010; Tedeschi, Fine, & Helgeson, 2010). These service dogs are thought to mitigate PTSD symptomology by instilling a sense of confidence, safety, and independence in the veteran on a day-to-day basis. Specific tasks can range from responding to and distracting a veteran from panic or emotional distress, “watching” their back in public, and waking them up from nightmares. PTSD service dogs may also alleviate anxious arousal/hypervigilance, avoidance, and feelings of isolation and detachment from others (Taylor, Edwards, & Pooley, 2013; Yeager & Irwin, 2012; Yount, Ritchie, Laurent, Chumley, & Olmert, 2013). As a result of the dog’s presence, individuals also report increased social confidence enabling them to leave their house, interact with friends and strangers, and reintegrate into society (Newton, 2014; Rubenstein, Debboun, & Burton, 2012; Stern et al., 2013; Yount et al., 2013).

Beyond anecdotal and largely retrospective reports, recent systematic reviews of the literature on Animal-Assisted Intervention (AAI) for PTSD reveal that there is a notable absence of peer-reviewed, empirical studies of the efficacy of service dogs for alleviating PTSD symptoms (Krause-Parello, Sarni, & Padden, 2016; O’Haire, Guérin, & Kirkham, 2015). Therefore, there is a critical need to evaluate proof-of-concept of the therapeutic efficacy of psychiatric service dogs for individuals with PTSD as this practice increases in its media attention and receives national financial and political attention from governmental organizations such as the U.S. Department of Veterans Affairs (Saunders et al., 2017). There is a further need for increased public awareness of the role and evidence-based outcomes of service dogs for PTSD.

The objective of this research was to empirically evaluate the effects of service dogs on standardized assessments of PTSD symptomology, depression, quality of life, and social and employment functioning in military members and veterans diagnosed with PTSD. To achieve this objective, we compared two groups: (1) individuals receiving usual care while on the waitlist to receive a service dog and (2) individuals receiving usual care plus the addition of a service dog. Our hypothesis was that participants with PTSD who have been placed with a service dog will show decreased PTSD symptom severity, decreased depression, increased quality of life, and greater overall social functioning compared to those receiving usual care while on the waitlist to receive a PTSD service dog.

Methods

Participants

Participants were recruited between November 2015 and February 2016 from a national sample of 304 individuals who applied and were approved to receive a trained PTSD service dog from an accredited service dog provider, K9s For Warriors. Inclusion criteria consisted of: (1) military service after September 11, 2001, (2) a clinician referral letter verifying a diagnosis of PTSD or meeting the clinical cutoff of 50 on the validated PTSD Checklist (PCL) (3) honorable discharge or current honorable service, (4) no substance abuse, (5) no conviction of any crime against animals, and (6) no more than two pet dogs currently in the home. A lack of prior history with substance and/or animal abuse were self-reported and confirmed by the organization via a background check. All participants had applied for and been approved to receive a service dog from K9s For Warriors. Approximately half of the sample was on the waitlist to receive a service dog and the other half had already received a service dog.

Usual care

Both participants on the waitlist and those with a service dog had unrestricted access to usual care during the course of the study. Unrestricted access indicates that no constraints were made on participant treatment usage. Participants were not directed towards any specific services, and were allowed to continue to receive intervention services and seek advice from medical professionals as they normally would. No statements were made to participants about their usual care, including continuing or altering treatment services. The frequency of PTSD treatment sessions and perceived improvements were recorded.

Waitlist group

Participants in the waitlist group had previously submitted an application to the service dog provider, K9s For Warriors. They had been on the waitlist between 2 months to 2.4 years (M = 0.64, SD = 0.36 years). Their application had been approved and they were currently waiting until their scheduled date to receive a service dog. Receipt of a service dog occurs in order of application.

Service dog group

Participants in the service dog group had been previously provided a PTSD service dog from K9s For Warriors. They had been paired with a service dog for between 1 month to 4 years (M = 1.64, SD = 1.07 years). The service dogs (predominantly Labrador Retrievers, Golden Retrievers, and Mixes) were primarily rescued from animal shelters and were selected based on a suite of characteristics ranging from physical size to temperamental demeanor. Participants placed with a service dog attended a three-week training class on site at K9s For Warriors headquarters in which they learned how to live with, care for, and maintain training with their future service dog before being sent home. Training classes consist of 6–8 recipients on average, in which all individuals live in dormitories on site and attend daily, scheduled activities to learn how to work with their service dogs both in public and private settings. Trained commands for each service dog ranged from basic obedience including sit, stay, down, and heel as well as a variety of commands specifically trained to mitigate PTSD symptoms. Examples of these trained tasks include alerting to anxiety or agitation to interrupt or prevent a panic attack, waking from nightmares, leaning against or standing in front of a veteran in social spaces to provide comfort and create personal space, retrieving and reminding to take medication, and allowing the veteran to physically brace on the dog for stabilization.

Procedure

The study protocol was approved by the Purdue University Human Research Protection Program Institutional Review Board (IRB Protocol 1504015973). No interactions occurred between the research team and the dogs during the course of the study; therefore a waiver was obtained from the Institutional Animal Care and Use Committee (IACUC).

The study consisted of both cross-sectional and longitudinal assessments. Cross-sectional assessments compared a single time point across the waitlist and service dog groups at the time of study, and thus directly compared current differences between those with and without a service dog. Longitudinal assessments consisted of previously collected PTSD assessments on file with the service dog provider, in addition to a current PTSD assessment as part of the study. The aim of the longitudinal assessments was to evaluate the trajectory of PTSD symptomology across five time points: (1) baseline (initial application to the service dog provider), (2) during the waitlist period, (3) immediately prior to service dog acquisition, (4) 3-weeks after service dog acquisition and (5) at follow-up once the service dog is in the home. The time points collected for each group are presented in Table 1.

Table 1.

Longitudinal assessments

Group
Time Point n Waitlist
(n = 66)		 Service Dog
(n = 75)
Waitlist
(1) Baseline (initial application for service dog) 60 X X
(2) During waitlist 66 X
(3) Before dog placement 33 X
Service Dog
(4) 3-weeks after dog placement 35 X
(5) Follow-up 74 X
Open in a new tab

Bolded X indicates cross-sectional comparison. All other time points were collected from records on file from the service dog provider.

For time points in the current study, all waitlisted or active recipients from the K9s For Warriors database were sent an initial study packet in the mail, which included information about the study protocol, participation materials, and $20 cash as remuneration for time spent reviewing materials. The mailing response rate was 46% (n = 141 of 304). Following voluntary informed consent, participants completed a battery of standardized, self-report assessments online (94%) or through the mail (6%). They also consented to allow research personnel to access their PTSD assessments on file with the service dog provider for longitudinal assessments. Upon completion of the study protocol, participants received an additional $20 in remuneration.

Measures

Usual Care

Participants’ usual care treatment was assessed with a subset of questions from the The American Legion Survey of Patient Healthcare Experiences (Greenberg, 2014). The questions asked if the participant currently received treatment for PTSD, TBI or MST as well as the frequency of treatment sessions and perceived level of improvement since receiving care on a scale of 1 to 10 (1=got worse, 5=no change, and 10=significantly better).

Posttraumatic Stress Disorder

The primary outcome measure of PTSD severity was assessed both cross-sectionally and longitudinally using the PTSD Checklist (PCL), a widely used 17-item scale based on the three DSM-IV symptom clusters of re-experiencing (subscale B) avoidance (subscale C) and arousal (subscale D) (Weathers, Litz, Herman, Huska, & Keane, 1993). A total score above 50 on a scale of 17 to 85 indicates a positive screening for PTSD for military personnel with a higher score indicating greater overall symptom severity (Forbes, Creamer, & Biddle, 2001). In addition, a change of 10 points is considered clinically meaningful. Cronbach’s α’s in the current sample were 0.85, 0.92, 0.92, 0.93 and 0.92 for the five assessment points, respectively.

Depression

Due to the multifaceted nature of depression, two outcome measures were enlisted to capture the breadth of self-reported depression characteristics. The Patient Health Questionnaire (PHQ-9) a 9-item tool for assessing depression and is commonly used for screening and diagnosis (α = 0.89) (Kroenke, Spitzer, & Williams, 2003). Cronbach’s α for the current sample was 0.95, indicating high reliability.

The Patient Reported Outcomes Measurement Information System (PROMIS) is a set of highly reliable, precise measures of physical, mental, and social well-being (Cella et al., 2010). The PROMIS Depression adult short-form 8-item scale was used with a higher score indicating greater depression. Cronbach’s α for the current sample was 0.84 indicating high reliability.

Quality of Life

The Veteran’s RAND 12 Item Health Survey (VR-12) is a health-related quality of life survey summarized into two scores, a physical component summary and a mental component summary (Iqbal et al., 2007). Higher scores indicate better overall health quality of life in either the mental or physical domain.

The Satisfaction With Life Scale (SWLS) is a 5-item instrument designed to measure judgments of satisfaction with one’s life (Diener, Emmons, Larsen, & Griffin, 1985). A higher score on the SWLS indicates higher life satisfaction. Cronbach’s α for the current sample was 0.85 indicating high reliability.

The Bradburn Scale of Psychological Wellbeing (BSPW) is a 5-item scale that assesses positive wellbeing (Bradburn, 1969). A higher score on the BSPW indicates higher positive well-being. Cronbach’s α for the current sample was 0.54 indicating moderate reliability.

The Connor Davidson Resilience Scale (CDRS) is a 25-item scale that measures resilience, or the capacity to change and cope with adversity (Connor & Davidson, 2003). A higher score on the CDRS indicates greater resilience. Cronbach’s α for the current sample was 0.92 indicating high reliability.

Social & Work Functioning

Three PROMIS scales were used to measure overall social functioning. The PROMIS Ability to Participate in Social Activities adult short form is an 8-item scale with higher scores indicating greater social participation. The PROMIS Social Isolation adult short form is an 8-item scale with higher scores indicating greater social isolation. The PROMIS Companionship adult short form is a 6-item scale with higher scores indicating a greater perceived level of companionship. Cronbach’s α’s for the current sample were 0.93, 0.91, and 0.93 for the three scales, respectively.

The Work Productivity and Activity Impairment Questionnaire: General Health V2.0 (WPAI) is a 6-item questionnaire that assesses the effect of an individual’s health problems on their ability to work and perform regular activities (Reilly, Zbrozek, & Dukes, 1993). It measures absenteeism and impairment at work as well as overall activity impairment as a result of one’s health.

Data Analysis

Data analyses were performed in two phases. First, demographic, military, and clinical characteristics of participants were compared using independent samples t-tests for continuous variables and chi-squared tests for categorical variables. Second, to examine differences in outcome measures based on condition (control vs service dog), we conducted a series of linear mixed effects models. Longitudinal data included participants followed during the waitlist period only (n = 66) as well as participants followed during both the waitlist and service dog periods (n = 75). For data with multiple time points per participant (i.e. PCL), we used hierarchical linear modeling (Raudenbush & Bryk, 2002) to examine differences in outcomes over time (within-subjects). For data with a single time point per participant (all other outcome measures), we compared the waitlist group to the service dog group using linear regression (Seber & Lee, 2012) to examine differences in outcome as a function of condition (between-subjects). In all models, sociodemographic variables (i.e. age, sex, marital status) were included as additional control covariates. Effect sizes are reported using Cohen’s d, with 0.20, 0.50, and 0.80 indicating small, medium, and large effect sizes, respectively (Cohen, 1992).

Results

Demographic characteristics

A total of 141 participants with PTSD completed the study, including 66 receiving usual care while on the waitlist and 75 receiving usual care while paired with a service dog. Demographic and military characteristics are displayed in Table 2. The sample was predominantly male (n = 110, 78.0%) with an average age of 37.1 years (S.D. = 8.3). Most participants had a significant other, spouse, or partner (n = 111, 78.7%) with an average of 3.1 people living in the household, including the participant (S.D. = 1.6). A subset of the sample (n = 55, 39.0%) required mobility aids. Approximately one-third of the sample had completed a college degree or higher (n = 47, 33.3%). The most common military branch was the Army (n = 93, 66.0%), with deployments to both Iraq (n = 90, 63.8%) and Afghanistan (n = 60, 42.6%) and the highest proportion of participants in the E4–E5 grade (n = 74, 52.5%). There were no significant differences between groups (service dog vs. control) on any demographic variable, except for marital status. A higher proportion of participants were married in the service dog group, compared to the control group [X2 (1, N = 141) = 8.23, p = .004], therefore marital status was included as a control variable in all models. At baseline (time of application to the service dog provider), there were no significant differences in PTSD Checklist scores (p = 0.732), indicating that both groups were similar in initial PTSD symptom severity.

Table 2.

Demographic and clinical characteristics of participants across groups

Group Group difference
Variable Waitlist
(n = 66)		 Service Dog
(n = 75)		 Total
(N = 141)		 t X2 p
Age, M (S.D.), y 37.3 (8.1) 37.0 (8.5) 37.1 (8.3) 0.206 0.837
Gender, n (%) Male 50 (75.8) 60 (80.0) 110 (78.0) 0.368 0.544
Marital Status, n (%) Single 21 (31.8) 9 (12.0) 30 (21.3) 8.232 0.004b
People in household, M (S.D.), # a 3.0 (1.5) 3.2 (1.8) 3.1 (1.6) −0.663 0.508
Using a Mobility Aid, n (%) 27 (40.9) 28 (37.3) 55 (39.0) 0.189 0.664
Traumatic Brain Injury (TBI) comorbidity n (%) 13 (19.7) 13 (17.3) 26 (18.4) 0.130 0.718
Education, n (%) 2.259 0.688
 High School or GED 4 (6.1) 8 (10.7) 12 (8.5)
 Some College 37 (56.1) 45 (60.0) 82 (58.2)
 Bachelor’s Degree 16 (24.2) 13 (17.3) 29 (20.6)
 Master’s Degree 8 (12.1) 7 (9.3) 15 (10.6)
 Advanced Graduate Work or PhD 1 (1.5) 2 (2.7) 3 (2.1)
Military Branch, n (%) 0.700 0.873
 Air Force 6 (9.1) 8 (10.7) 14 (9.9)
 Army 45 (68.2) 48 (64.0) 93 (66.0)
 Marines 8 (12.1) 8 (10.7) 16 (11.3)
 Navy 7 (10.6) 11 (14.7) 18 (12.8)
Military Grade, n (%) 4.289 0.368
 E1 – E3 3 (4.5) 6 (8.0) 9 (6.4)
 E4 – E5 35 (53.0) 39 (52.0) 74 (52.5)
 E6 – E7 17 (25.8) 20 (26.7) 37 (26.2)
 E8 – E9 3 (4.5) 6 (8.0) 9 (6.4)
 Officer 8 (12.1) 3 (4.0) 11 (7.8)
 Unknown 0 (0.0) 1 (1.3) 1 (0.7)
Deployment, n (%)
 Iraq 43 (65.2) 47 (62.7) 90 (63.8) 0.759 0.094 0.759
 Afghanistan 27 (40.9) 33 (44.0) 60 (42.6) 0.711 0.137 0.711
PTSD Checklist, M (S.D.), Baseline c 70.2 (8.7) 69.4 (8.8) 69.7 (8.7) 0.344 0.732
 Re-experiencing (B) Subscale, M (S.D.) 20.3 (3.5) 19.6 (3.3) 19.8 (3.7) 0.729 0.469
 Avoidance (C) Subscale, M (S.D.) 28.0 (3.9) 28.5 (3.9) 28.3 (3.9) −0.493 0.624
 Arousal (D) Subscale, M (S.D.) 22.0 (3.4) 21.3 (3.2) 21.6 (3.3) 0.752 0.455
Open in a new tab

M, mean; S.D., standard deviation; y, years; n, partial sample size; N, total sample size; %, percentage of participants; PTSD, posttraumatic stress disorder

a

Including the veteran

b

Marital status was included as a control variable in all models

c

Baseline score at time of initial application to service dog provider

Usual care

The American Legion Treatment Survey was used to ascertain usual care treatment participation across groups, which are displayed in Table 3. These include non-service dog PTSD treatments. There were no significant differences between groups in the number of participants currently receiving treatment (p = 0.940, d = 0.07) nor in how frequently they received treatment sessions per year (p = .482, d = 0.05), indicating that both groups were roughly equivalent in their dosage of usual care services. However, compared to the waitlist group, participants with service dogs reported a higher overall level of perceived improvement from their treatment with a medium effect size (p = .007, d = 0.55), indicating that those with service dogs perceive greater improvement from the same dosage of usual care treatment services.

Table 3.

Usual care PTSD treatment participation across groups

Group Group difference
Measure Waitlist
(n = 66)		 Service Dog
(n = 75)		 b t d
Receiving PTSD treatment, n (%) 53 (80.3) 58 (77.3) 0.04 0.08 0.07
Treatment sessions per year, M (S.D.) 39.0 (38.7) 41.9 (78.0) 7.87 0.71 0.05
Perceived level of improvement since receiving care, M (S.D.) 5.0 (1.9) 6.0 (1.7) 0.98** 2.76 0.55
Open in a new tab

PTSD, posttraumatic stress disorder; M, mean; S.D., standard deviation; b, standardized regression coefficient (reference category: service dog), d, Cohen’s d effect size; CI, confidence interval;

**

, p < .01

a

Perceived level of improvement since receiving care, M (SD)

Service dog outcomes

The longitudinal assessment compared PTSD symptomology within individuals, including up to three time points while on the waitlist and up to two time points with a service dog. Results of longitudinal PCL scores are reported in Table 4. Compared to baseline (initial application to the service dog provider), there were no significant differences on the PCL at any point during the waitlist period (during waitlist: p = .202, end of waitlist: p = .504); however, there were significant reductions on the PCL at both points during the service dog period with large effect sizes (after 3 weeks: p < .001, d = −2.11, follow-up: p < .001, d = −1.03). Estimated reductions from baseline were between 11.54 and 21.36 points on average, which is larger than the standard cutoff of 10 points indicating a clinically meaningful change in PTSD symptomology.

Table 4.

Longitudinal comparison of PTSD Checklist (PCL) scores over time within participants

Total PCL Re-experiencing
(subscale B)		 Avoidance
(subscale C)		 Arousal
(subscale D)
Condition n M (S.D.) b d M (S.D.) b d M (S.D.) b d M (S.D.) b d
Waitlist
 Baseline (application for dog) 60 69.7 (8.7) – 19.8 (3.4) – 28.3 (3.9) – 21.6 (3.3) –
 During waitlist a 66 66.3 (11.7) −2.60 −0.32 18.4 (4.6) −1.19 −0.35 27.1 (5.0) −0.10 −0.27 20.9 (3.4) −0.43 −0.34
 Before dog placement b 33 70.7 (10.7) 1.65 0.11 19.7 (3.6) 0.01 −0.03 29.3 (4.8) 1.26 0.23 21.7 (3.4) 0.39 0.04
Service Dog
 3-weeks after dog placement b 35 47.9 (11.7) −21.36*** −2.11 14.6 (4.1) −5.17*** −1.38 18.2 (5.0) −9.94*** −2.25 15.1 (4.3) −6.25*** −1.43
 Follow-up a 74 58.2 (13.1) −11.54*** −1.03 16.5 (4.3) −3.28*** −0.85 23.2 (6.2) −5.17*** −0.98 18.5 (4.3) −3.08*** −0.90
Open in a new tab

PTSD, posttraumatic stress disorder; n, sample size; M, mean; S.D., standard deviation; b, unstandardized coefficient (reference category: baseline), d, effect size,

*

, p < .05;

**

, p < .01;

***

, p < .001

a

Data from cross-sectional between group comparison at a single time point.

b

Training consisted of a 3-week period on site at the service dog provider.

The cross-sectional, single time point assessment compared functioning between individuals receiving usual care while on the waitlist (control) to those receiving usual care in addition to a service dog (treatment). Results of linear regression models are reported in Table 5. Compared to the control group, participants with a service dog demonstrated significantly lower scores for PTSD symptomology with a medium effect size on the PCL (p < .001, d = −0.66). There was no significant correlation between baseline PCL score and change over time (to the cross-sectional survey time point) in the service dog group (r = −.193, p = .238), indicating that initial PTSD symptom severity was not associated with service dog outcomes on the PCL.

Table 5.

Comparison of outcomes between groups at a cross-sectional time point

Group Group difference
Measure Waitlist (n = 66)
M (S.D.)		 Service Dog (n = 75)
M (S.D.)		 b t d
Depression
 PROMIS Depression SF 8a 28.9 (7.4) 22.3 (7.2) −7.07*** −5.68 −0.91
 PHQ-9 17.9 (5.3) 14.0 (5.4) −4.33*** −4.62 −0.73
Quality of Life
 VR-12 Mental Health 24.4 (9.7) 30.9 (10.1) 6.84*** 3.87 0.66
 VR-12 Physical Health 37.1 (12.3) 36.8 (10.9) −0.24 −0.12 −0.02
 BSPW −2.7 (2.0) −0.9 (2.5) 1.86*** 4.72 0.81
 SWLS 15.0 (5.9) 18.8 (7.9) 3.58** 3.00 0.55
 CDRS 18.5 (7.3) 22.8 (8.5) 4.89*** 3.67 0.54
Social
 PROMIS Ability to Participate in Social Activities SF 8a 16.2 (5.7) 20.8 (6.9) 5.11*** 4.83 0.73
 PROMIS Social Isolation SF 8a 30.6 (6.3) 26.7 (6.8) −4.41*** −3.95 −0.60
 PROMIS Companionship SF 6a 19.0 (5.4) 22.1 (6.5) 2.00* 2.05 0.52
Work
 WPAI – Employed, n (%) 18 (24.7%) 22 (33.3%) 0.30 0.76 −0.19
 WPAI – Absenteeisma 27.6 (35.2) 5.0 (8.4) −22.77* −2.46 −0.89
 WPAI – Impairment at Work (Health)a 52.7 (30.9) 44.4 (25.0) −7.33 −0.76 −0.29
 WPAI – Impairment at Work (Overall)a 64.4 (29.7) 44.8 (27.4) −19.99† −2.03 −0.69
 WPAI – Activity Impairmenta 62.6 (27.7) 56.3 (26.9) −9.43* −1.99 −0.23
Open in a new tab

M, mean; S.D., standard deviation; b, standardized regression coefficient (reference category: service dog), d, Cohen’s d effect size; CI, confidence interval;

†

p < .10;

*

p < .05;

**

p < .01;

***

p < .001;

PCL, PTSD Checklist; PROMIS, Patient-Reported Outcomes Measurement Information System; SF, short form; PHQ-9, Patient Health Questionnaire 9; VR-12 Mental Health, Veteran’s Rand 12 item Health Survey-Mental Health Component; VR-12 Physical Health, Veteran’s Rand 12 item Health Survey-Physical Health Component; BSPW, Bradburn Scale of Psychological Wellbeing; SWLS, Satisfaction with Life Scale; CDRS, Connor-Davidson Resilience Scale; WPAI, Work Productivity and Activity Impairment Questionnaire: General Health Problem V2.0

a

Among veterans who are employed

Participants with service dogs exhibited significantly lower depression symptomology with a large effect size on the PROMIS Depression (p < .001, d = −0.91) and a medium effect size on the PHQ-9 (p < .001, d = −0.74). Quality of life was higher among those with service dogs, with medium to large effect sizes on the VR-12 Mental (p < .001, d = 0.66), BSPW (p < .001, d = 0.81), SWLS (p = .003, d = 0.59), and CDRS (p < .001, d = 0.55). No significant differences were reported on the VR-12 Physical (p = .908, d = −0.03). Compared to the control group, participants with a service dog reported significantly higher social functioning with medium effect sizes, including a greater ability to participate in social activities (PROMIS Social Activities: p < .001, d = 0.70), lower social isolation (PROMIS Social Isolation: p < .001, d = −0.63), and higher perceived companionship (PROMIS Companionship: p = .043, d = 0.52). There were no significant differences in the proportion of individuals who were employed between groups (WPAI: p = .451, d = −0.20); however, for those who were working, individuals with service dogs reported a lower proportion of work missed due to health with a large effect size (p = .019, d = −0.89) and a lower rate of activity impairment with a small effect size (p = .049, d = −0.27), but no significant differences in their level of impairment while at work overall (p = .051, d = −0.69) or due to health (p = .453, d = −0.29).

Discussion

The purpose of this study was to examine the preliminary efficacy of trained service dogs for military members and veterans with PTSD, compared to a usual care waitlisted control group. The results indicated that compared to usual care alone, the provision of trained service dogs was associated with clinically significant reductions in PTSD symptoms on the PTSD Checklist. However, average scores were not lower than the diagnostic cutoff of 50 on the PTSD Checklist, indicating that in their current form, service dogs do not appear to be associated with a loss of diagnosis. This research presents proof-of-concept that in combination with usual care, service dogs may reduce perceived PTSD symptoms among military members and veterans. These findings offer support for initial efficacy, but require further research to evaluate their integration with evidence-based treatments.

Changes in PTSD symptomology may be due to the emerging body of evidence suggesting that the presence of animals influences socio-emotional functioning in non-military PTSD populations (Bert et al., 2016; Hart, 2006; Wells, 2009). These findings map roughly onto the diagnostic criterion for PTSD related to negative alterations in cognition and mood. Studies have demonstrated that the presence of a dog can reduce feelings of social estrangement (Allen & Blascovich, 1996; Rintala, Sachs-Ericsson, & Hart, 2002; Wood, Giles-Corti, & Bulsara, 2005), motivate social participation (Barak, Savorai, Mavashev, & Beni, 2001; Fairman & Huebner, 2001; Taylor et al., 2013), produce positive emotions (Collins et al., 2006) and reduce negative emotions (Souter & Miller, 2007). It is possible that these effects are elicited by the addition of a service dog for individuals with PTSD, and that they either indirectly or directly influence pathways to reductions PTSD symptomology. Indeed, secondary outcomes revealed that relative to usual care alone, individuals with a service dog exhibited significant differences with medium to large effect sizes in some of these domains. Specifically, those with service dogs showed differences with respect to depression (lower symptomology), quality of life (increased mental, but not physical, quality of life, increased psychological wellbeing, life satisfaction, and resilience), social functioning (increased ability to participate in social activities, lower social isolation, greater feelings of companionship), and some differences with respect to work functioning (no differences in employment level or impairment at work, but lower absenteeism and activity impairment due to health).

Historically, service dogs have been partnered with individuals with physical or ambulatory disabilities by assisting with mobility tasks (Winkle, Crowe, & Hendrix, 2012). The provision of service dogs to address psychosocial needs has emerged in recent years (Tedeschi et al., 2010). The efficacy of service dogs for participants with PTSD in this sample appears to be tailored to mental health, rather than physical health outcomes. For example, on the VR-12 quality of life measure, individuals with a service dog scored significantly higher on the mental health component of the measure, but the physical health component did not differ between groups. Similarly, on the WPAI work productivity measure, the overall health impairment at work component was significantly lower among those with service dogs, but the physical impairment at work component was not different between groups. These characteristics suggest that compared to usual care alone, trained service dogs for PTSD are related to primarily psychosocial differences rather than purely physical differences.

Concern has been expressed that some individuals may seek animal-assisted interventions in place of evidence-based treatments, putting them at risk of not receiving effective services (Anestis, Anestis, Zawilinski, Hopkins, & Lilienfeld, 2014). The results of this study contradict this assertion; they suggest that participants with service dogs are receiving similar levels of PTSD treatment (usual care) to those on the waitlist (>75% in both groups). Thus in the current sample, participants did not employ service dogs to substitute treatment as usual, but instead added service dogs to complement treatment as usual. The only difference was that participants with service dogs perceived a higher level of improvement (20% higher on average) from the same dosage of usual care treatment. Though significantly higher than the waitlist group, the service dog group perceived only slightly more than “no change” from their usual care treatments. It is unclear why participants with service dogs would perceive more improvement from the same level of treatment; however, it may be due to co-occurring increased feelings of resilience and ability to participate in social activities, which could create a more engaging space for the implementation of evidence-based practices.

The findings from this preliminary study also suggest that the outcomes from service dogs are comparable to those of evidence-based practices for PTSD. The results indicate that on average, the provision of a service dog in combination with treatment as usual contributed to a clinically significant reduction in PTSD symptoms, but not below a conservative diagnostic cutoff of 50 on the self-reported PCL. These findings mirror a review of randomized clinical trials of evidence-based treatments for PTSD, where mean post-treatment scores also remained at or above the clinical cutoff for PTSD (e.g. (Monson et al., 2006; Steenkamp, Litz, Hoge, & Marmar, 2015). Within-group treatment effect sizes from this research are also similar to frequently studied psychotherapies for military PTSD (Steenkamp et al., 2015; Watts et al., 2013). For example, pre-post effect sizes for cognitive processing therapy (CPT) and prolonged exposure (PE; Cohen’s d range = 0.78–1.10) are comparable to service dogs in the current study at the follow-up time point in the longitudinal PCL analysis (Cohen’s d = 1.03; Steenkamp et al., 2015).

Limitations

Outcomes from the current study should be interpreted with consideration of some important limitations. First, the control condition was usual care, which can include participants receiving no treatment at all. It is unknown what types of treatments participants were receiving; thus although they received the same number of sessions per year, the types of sessions may have varied across groups. The control condition also does not account for the potential effects of non-specific treatment factors such as attention or novelty. This limitation is particularly salient with respect to the first three weeks of the treatment period, which include training on site at the service dog provider with a small cohort of fellow service dog recipients. The active components and effects of this unique time period are a critical area for further investigation as the in-person training session may act as a form of treatment in itself with or without the service dog component being present (Yount et al., 2013). However, following the introduction of the service dog into the home, there are minimal non-specific treatment effects related to attention from a therapist or treatment group, given that the only component of the intervention is the dog itself.

The goal of this study was to conduct an ecologically valid preliminary efficacy study to determine the effects of a service dog compared to unrestricted access to usual care, which included evidence-based treatments. Based on the current results, further studies can be developed to enlist an active comparison that accounts for these limitations as well as possible placebo effects (Furukawa et al., 2014). The specific roles and usage of trained dog commands should also be investigated to empirically define the treatment and evaluate fidelity and best practices.

Second, allocation to treatment group was not randomized. Results may have been due to natural maturational changes over time, rather than the service dog. Given the multi-year long waitlists associated with service dogs for PTSD, it was not possible for ethical reasons to change order on the waitlist to randomize for this study. Third, the sample was self-selected as a group of individuals who had a demonstrated interest in obtaining a service dog. Recruitment consisted of contacting individuals who had already applied for or received a service dog. Thus, the results should be interpreted as generalizable only to those who are amenable to service dogs. At a minimum, our findings provide initial support that service dogs do not seem to have aversive outcomes for those who are motivated to get them.

Finally, the results include standardized self-report, which may be subject to expectancy biases (Cook, 2010). It is possible that baseline measurements were inflated at the time of applying to the service dog provider to justify the need for a service dog. Inflation of symptoms represents an interesting shift from the documented denial and minimization among many military personnel, who underrate their PTSD symptoms to avoid a diagnosis (Davidson & Connor, 1999). If military personnel are willing to exaggerate symptoms to receive a diagnosis and be paired with a service dog, this may evidence the perceived value of the service dog despite the associated stigma. To address biases in self-reported symptomology, the incorporation of a validated diagnosis of PTSD from a private or community health provider was enlisted to authenticate baseline PCL values; however, a more objective measure of PTSD severity such as the clinician-administered PTSD scale (CAPS; Blake et al., 1995) would have been beneficial.

Future studies should enlist physiological and blinded assessments to obtain additional objective indicators of change over time and between groups rather than relying on standardized self-report alone. Individual differences in evidence-based treatment receptivity and concurrent diagnoses should also be explored. Further studies should incorporate a comprehensive intake to define usage of different types of usual care. This will enable evaluation of how the provision of a service dog may change participation in usual care. It will also foster evaluation of the best ways to incorporate service dogs as an adjunct to evidence-based treatment.

Conclusions

This pragmatic, longitudinal effectiveness trial provides initial evidence that compared to usual care alone, military members and veterans with trained service dogs show lower PTSD symptomology, reduced depression, and increased social participation. Individual differences influencing short term and long term efficacy remain to be tested. Ongoing research is needed to determine the most effective ways to incorporate service dogs into evidence-based usual care as well as how to enhance service dog best practices to achieve maximal clinical change.

What is the public health significance of this article?

Despite anecdotal accounts of the benefits of service dogs for military members and veterans with PTSD, limited empirical evidence exists to document their efficacy. This proof-of-concept study suggests that the addition of trained psychiatric service dogs to usual care may improve PTSD symptomology, but not below the level of clinical diagnosis, and contribute to a greater quality of life and improved social functioning. In their current form, service dogs may confer benefits as a complementary or integrative treatment option among military members and veterans with PTSD.

Acknowledgments

We acknowledge the participation of our collaborators at K9s For Warriors for their assistance in completing the study.

This study was funded by the Human-Animal Bond Research Initiative (HABRI) Foundation (Grant D15HA-031) and Bayer Animal Health. This publication was made possible with support from Grant Numbers KL2TR001106 and UL1TR001108 (A. Shekhar, PI) from the National Institutes of Health, National Center for Advancing Translational Sciences, Clinical and Translational Sciences Award.

References

  1. Allen K, Blascovich J. The value of service dogs for people with severe ambulatory disabilities. A randomized controlled trial. JAMA. 1996;275(13):1001–1006. [PubMed] [Google Scholar]
  2. American Psychiatric Association. Diagnostic and statistical manual of mental disorders: DSM-V. 5th. Washington, DC: Author; 2013. [Google Scholar]
  3. Anestis MD, Anestis JC, Zawilinski LL, Hopkins TA, Lilienfeld SO. Equine-related treatments for mental disorders lack empirical support: A systematic review of empirical investigations. Journal of clinical psychology. 2014;70(12):1115–1132. doi: 10.1002/jclp.22113. [DOI] [PubMed] [Google Scholar]
  4. Barak Y, Savorai O, Mavashev S, Beni A. Animal-assisted therapy for elderly schizophrenic patients: A one-year controlled trial. American Journal of Geriatric Psychiatry. 2001;9(4):439–442. [PubMed] [Google Scholar]
  5. Bert F, Gualano MR, Camussi E, Pieve G, Voglino G, Siliquini R. Animal assisted intervention: A systematic review of benefits and risks. European Journal of Integrative Medicine. 2016 doi: 10.1016/j.eujim.2016.05.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Blake DD, Weathers FW, Nagy LM, Kaloupek DG, Gusman FD, Charney DS, Keane TM. The development of a clinician-administered PTSD scale. Journal of Traumatic Stress. 1995;8(1):75–90. doi: 10.1007/BF02105408. [DOI] [PubMed] [Google Scholar]
  7. Bomyea J, Lang AJ. Emerging interventions for PTSD: future directions for clinical care and research. Neuropharmacology. 2012;62(2):607–616. doi: 10.1016/j.neuropharm.2011.05.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bradburn NM. The structure of psychological well-being. Chicago: Adline; 1969. [Google Scholar]
  9. Cella D, Riley W, Stone A, Rothrock N, Reeve B, Yount S, Choi S. The Patient-Reported Outcomes Measurement Information System (PROMIS) developed and tested its first wave of adult self-reported health outcome item banks: 2005–2008. Journal of clinical epidemiology. 2010;63(11):1179–1194. doi: 10.1016/j.jclinepi.2010.04.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Cohen J. A power primer. Psychological bulletin. 1992;112(1):155. doi: 10.1037//0033-2909.112.1.155. [DOI] [PubMed] [Google Scholar]
  11. Collins DM, Fitzgerald SG, Sachs-Ericsson N, Scherer M, Cooper RA, Boninger ML. Psychosocial well-being and community participation of service dog partners. Disabil Rehabil Assist Technol. 2006;1(1–2):41–48. doi: 10.1080/09638280500167183. [DOI] [PubMed] [Google Scholar]
  12. Connor KM, Davidson JR. Development of a new resilience scale: The Connor-Davidson resilience scale (CD-RISC) Depression and Anxiety. 2003;18(2):76–82. doi: 10.1002/da.10113. [DOI] [PubMed] [Google Scholar]
  13. Cook C. Mode of administration bias. Journal of Manual & Manipulative Therapy. 2010;18(2):61–63. doi: 10.1179/106698110X12640740712617. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Cukor J, Spitalnick J, Difede J, Rizzo A, Rothbaum BO. Emerging treatments for PTSD. Clinical psychology review. 2009;29(8):715–726. doi: 10.1016/j.cpr.2009.09.001. [DOI] [PubMed] [Google Scholar]
  15. Davidson JRT, Connor KM. Management of posttraumatic stress disorder: Diagnostic and therapeutic issues. The Journal of Clinical Psychiatry. 1999;60(Suppl 18):33–38. [PubMed] [Google Scholar]
  16. Diener E, Emmons RA, Larsen RJ, Griffin S. The satisfaction with life scale. Journal of Personality Assessment. 1985;49(1):71–75. doi: 10.1207/s15327752jpa4901_13. [DOI] [PubMed] [Google Scholar]
  17. Fairman SK, Huebner RA. Service dogs: A compensatory resource to improve function. Occupational Therapy in Health Care. 2001;13(2):41–52. doi: 10.1080/J003v13n02_03. [DOI] [PubMed] [Google Scholar]
  18. Forbes D, Creamer M, Biddle D. The validity of the PTSD checklist as a measure of symptomatic change in combat-related PTSD. Behaviour research and therapy. 2001;39(8):977–986. doi: 10.1016/s0005-7967(00)00084-x. [DOI] [PubMed] [Google Scholar]
  19. Furukawa T, Noma H, Caldwell D, Honyashiki M, Shinohara K, Imai H, Churchill R. Waiting list may be a nocebo condition in psychotherapy trials: a contribution from network meta-analysis. Acta Psychiatrica Scandinavica. 2014;130(3):181–192. doi: 10.1111/acps.12275. [DOI] [PubMed] [Google Scholar]
  20. Greenberg J. The American Legion Survey of Patient Healthcare Experiences. Paper presented at the Advancing Care and Treatments for Veterans with TBI and PTSD; Washington, D.C.. 2014. [Google Scholar]
  21. Grieger TA, Cozza SJ, Ursano RJ, Hoge C, Martinez PE, APRN-PMH B, Wain HJ. Posttraumatic stress disorder and depression in battle-injured soldiers. American Journal of Psychiatry. 2006 doi: 10.1176/ajp.2006.163.10.1777. [DOI] [PubMed] [Google Scholar]
  22. Hart LA. Community context and psychosocial benefits of animal companionship. In: Fine AH, editor. Handbook on animal-assisted therapy: Theoretical foundations and guidelines for practice. 2nd. San Diego, CA: Academic Press; 2006. pp. 73–94. [Google Scholar]
  23. Hoge CW, Castro CA, Messer SC, McGurk D, Cotting DI, Koffman RL. Combat duty in Iraq and Afghanistan, mental health problems, and barriers to care. New England Journal of Medicine. 2004;351(1):13–22. doi: 10.1056/NEJMoa040603. [DOI] [PubMed] [Google Scholar]
  24. Hoge CW, Grossman S, Auchterlonie J, Riviere L, Milliken C, Wilk J. PTSD treatment for soldiers after combat deployment: low utilization of mental health care and reasons for dropout. Psychiatr Serv. 2014;65(8):997–1004. doi: 10.1176/appi.ps.201300307. [DOI] [PubMed] [Google Scholar]
  25. Iqbal SU, Rogers W, Selim A, Qian S, Lee A, Ren X, Kazis L. The Veterans RAND 12 Item Health Survey (VR-12): what it is and how it is used. CHQOERs VA Medical Center Bedford, MA: CAPP Boston University School of Public Health. 2007:1–12. [Google Scholar]
  26. Jacobson IG, Ryan MA, Hooper TI, Smith TC, Amoroso PJ, Boyko EJ, Bell NS. Alcohol use and alcohol-related problems before and after military combat deployment. JAMA. 2008;300(6):663–675. doi: 10.1001/jama.300.6.663. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kang HK, Natelson BH, Mahan CM, Lee KY, Murphy FM. Post-traumatic stress disorder and chronic fatigue syndrome-like illness among Gulf War veterans: a population-based survey of 30,000 veterans. American journal of epidemiology. 2003;157(2):141–148. doi: 10.1093/aje/kwf187. [DOI] [PubMed] [Google Scholar]
  28. Kemp J, Bossarte R. Suicide data report: 2012. Department of Veterans Affairs, Mental Health Services, Suicide Prevention Program Washington; DC: 2013. [Google Scholar]
  29. Krause-Parello CA, Sarni S, Padden E. Nurse Education Today. Advance online publication; 2016. Military veterans and canine assistance for post-traumatic stress disorder: A narrative review of the literature. [DOI] [PubMed] [Google Scholar]
  30. Kroenke K, Spitzer RL, Williams JB. The Patient Health Questionnaire-2: validity of a two-item depression screener. Medical care. 2003;41(11):1284–1292. doi: 10.1097/01.MLR.0000093487.78664.3C. [DOI] [PubMed] [Google Scholar]
  31. Kruger KA, Serpell JA. Animal-assisted interventions in mental health: Definitions and theoretical foundations. In: Fine AH, editor. Handbook on animal-assisted therapy: Theoretical foundations and guidelines for practice. 3rd. San Diego: Academic Press; 2010. pp. 33–48. [Google Scholar]
  32. Kulka RA, Schlenger WE, Fairbank JA, Hough RL, Jordan BK, Marmar CR, Weiss DS. Trauma and the Vietnam war generation: Report of findings from the National Vietnam Veterans Readjustment Study. New York, NY: Brunner/Mazel; 1990. [Google Scholar]
  33. Monson CM, Schnurr PP, Resick PA, Friedman MJ, Young-Xu Y, Stevens SP. Cognitive processing therapy for veterans with military-related posttraumatic stress disorder. Journal of Consulting and Clinical Psychology. 2006;74(5):898. doi: 10.1037/0022-006X.74.5.898. [DOI] [PubMed] [Google Scholar]
  34. Mott JM, Mondragon S, Hundt NE, Beason-Smith M, Grady RH, Teng EJ. Characteristics of U.S. Veterans Who Begin and Complete Prolonged Exposure and Cognitive Processing Therapy for PTSD. Journal of Traumatic Stress. 2014;27(3):265–273. doi: 10.1002/jts.21927. [DOI] [PubMed] [Google Scholar]
  35. Newton R. Exploring the Experiences of Living With Psychiatric Service Dogs for Veterans with Posttraumatic Stress Disorder. 1557029 MA. Adler School of Professional Psychology 2014 [Google Scholar]
  36. O’Haire ME, Guérin NA, Kirkham AC. Animal-assisted intervention for trauma: A systematic literature review. Frontiers in psychology. 2015;6 doi: 10.3389/fpsyg.2015.01121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Ouimette P, Vogt D, Wade M, Tirone V, Greenbaum MA, Kimerling R, Rosen CS. Perceived barriers to care among veterans health administration patients with posttraumatic stress disorder. Psychological Services. 2011;8(3):212. [Google Scholar]
  38. Pietrzak RH, Johnson DC, Goldstein MB, Malley JC, Southwick SM. Perceived stigma and barriers to mental health care utilization among OEF-OIF veterans. Psychiatric Services. 2015 doi: 10.1176/ps.2009.60.8.1118. [DOI] [PubMed] [Google Scholar]
  39. Raudenbush SW, Bryk AS. Hierarchical linear models: Applications and data analysis methods. Vol. 1. Sage; 2002. [Google Scholar]
  40. Reilly MC, Zbrozek AS, Dukes EM. The validity and reproducibility of a work productivity and activity impairment instrument. Pharmacoeconomics. 1993;4(5):353–365. doi: 10.2165/00019053-199304050-00006. [DOI] [PubMed] [Google Scholar]
  41. Rintala D, Sachs-Ericsson N, Hart K. The effects of service dogs on the lives of persons with mobility impairments: A pre-post study design. SCI Psychosocial Process. 2002;15(2):70–82. [Google Scholar]
  42. Rubenstein DA, Debboun M, Burton R. Canine-assisted therapy in military medicine. Perspectives: commander’s introduction. US Army Medical Department journal. 2012:1–4. [PubMed] [Google Scholar]
  43. Saunders GH, Biswas K, Serpi T, McGovern S, Groer S, Stock EM, Abrams T. Design and challenges for a randomized, multi-site clinical trial comparing the use of service dogs and emotional support dogs in Veterans with post-traumatic stress disorder (PTSD) Contemporary Clinical Trials. 2017 doi: 10.1016/j.cct.2017.08.017. [DOI] [PubMed] [Google Scholar]
  44. Sayer NA, Friedemann-Sanchez G, Spoont M, Murdoch M, Parker LE, Chiros C, Rosenheck R. A qualitative study of determinants of PTSD treatment initiation in veterans. Psychiatry. 2009;72(3):238–255. doi: 10.1521/psyc.2009.72.3.238. [DOI] [PubMed] [Google Scholar]
  45. Schottenbauer MA, Glass CR, Arnkoff DB, Tendick V, Gray SH. Nonresponse and dropout rates in outcome studies on PTSD: review and methodological considerations. Psychiatry. 2008;71(2):134–168. doi: 10.1521/psyc.2008.71.2.134. [DOI] [PubMed] [Google Scholar]
  46. Seber GA, Lee AJ. Linear regression analysis. Vol. 936. John Wiley & Sons; 2012. [Google Scholar]
  47. Souter MA, Miller MD. Do animal-assisted activities effectively treat depression? A meta-analysis. Anthrozoös. 2007;20(2):167–180. [Google Scholar]
  48. Steenkamp MM, Litz BT, Hoge CW, Marmar CR. Psychotherapy for military-related PTSD: a review of randomized clinical trials. JAMA. 2015;314(5):489–500. doi: 10.1001/jama.2015.8370. [DOI] [PubMed] [Google Scholar]
  49. Stern SL, Donahue DA, Allison S, Hatch JP, Lancaster CL, Benson TA, Peterson AL. Potential benefits of canine companionship for military veterans with posttraumatic stress disorder (PTSD) Society and Animals. 2013;21(6):568–581. doi: 10.1163/15685306-12341286. [DOI] [Google Scholar]
  50. Tanielian T, Jaycox LH. Invisible wounds of war. Santa Monica, CA: Rand Center; 2008. Retrieved on July, 11, 2008. [Google Scholar]
  51. Taylor MF, Edwards ME, Pooley JA. “Nudging them back to reality”: Toward a growing public acceptance of the role dogs fulfill in ameliorating contemporary veterans’ PTSD symptoms. Anthrozoös. 2013;26(4):593–611. [Google Scholar]
  52. Tedeschi P, Fine AH, Helgeson JI. Assistance animals: Their evolving role in psychiatric service applications. In: Fine AH, editor. Handbook on animal-assisted therapy: Theoretical foundations and guidelines for practice. 3rd. San Diego, CA: Academic Press; 2010. pp. 421–438. [Google Scholar]
  53. Trimble MR. Post-traumatic stress disorder: History of a concept. Trauma and its wake. 1985;1:5–14. [Google Scholar]
  54. Watts BV, Schnurr PP, Mayo L, Young-Xu Y, Weeks WB, Friedman MJ. Meta-analysis of the efficacy of treatments for posttraumatic stress disorder. The Journal of Clinical Psychiatry. 2013;74(6):e541–550. doi: 10.4088/JCP.12r08225. [DOI] [PubMed] [Google Scholar]
  55. Weathers FW, Litz BT, Herman DS, Huska JA, Keane TM. The PTSD Checklist (PCL): Reliability, validity, and diagnostic utility. Paper presented at the 9th Annual Meeting of the International Society for Traumatic Stress Studies; San Antonio, TX. 1993. [Google Scholar]
  56. Wells DL. The effects of animals on human health and well-being. Journal of Social Issues. 2009;65(3):523–543. [Google Scholar]
  57. Winkle M, Crowe TK, Hendrix I. Service dogs and people with physical disabilities partnerships: A systematic review. Occupational therapy international. 2012;19(1):54–66. doi: 10.1002/oti.323. [DOI] [PubMed] [Google Scholar]
  58. Wood L, Giles-Corti B, Bulsara M. The pet connection: Pets as a conduit for social capital? Social Science & Medicine. 2005;61(6):1159–1173. doi: 10.1016/j.socscimed.2005.01.017. [DOI] [PubMed] [Google Scholar]
  59. Yeager AF, Irwin J. Rehabilitative canine interactions at the Walter Reed National Military Medical Center. US Army Medical Department Journal. 2012:57–61. [PubMed] [Google Scholar]
  60. Yount RA, Ritchie EC, Laurent MS, Chumley P, Olmert MD. The role of service dog training in the treatment of combat-related PTSD. Psychiatric Annals. 2013;43(6):292–295. [Google Scholar]

RESOURCES