We adopted a mixed methods approach to offer breadth and depth of understanding beyond what qualitative or quantitative methods would allow alone (39). Specifically, our mixed method design was an explanatory unidirectional approach (40) where questionnaire data merged with qualitative findings from thematic analysis and in-depth case study analysis. Interview data was reported using a qualitative descriptive design with narratives (41). The study was approved by the Indiana University Institutional Review Board and VA Research and Development Committee.

This study included a sample of CPs participating in a 2-year longitudinal study that examined community reintegration among military Veterans with an invisible injury (42), which includes clinical diagnosis of post-traumatic stress disorder, depression, anxiety, traumatic brain injury, or other mental or cognitive health condition and their care partners. As reported elsewhere, 91% of these Veteran participants had a disability rating (43). A CP is defined by the Veteran as someone who supports him/her in an area considered important to reintegration (family/social life, school, work, rehabilitation, etc.), typically a family member, partner, friend, or neighbor (33). Following the baseline visit, Veterans were encouraged to nominate an individual who currently supports their adjustment to civilian life. Nominated individuals were contacted via phone to describe the study and inquire about their interest in study participation. Of the 75 Veterans, 48 nominated a CP, which led to a convenience sample of 36 CPs enrolled in the study.

Before collecting any data, a member of the research team discussed the study aims with CPs and obtained consent and HIPAA authorization. Data collection included a mix of quantitative and qualitative data. The qualitative data collection involved semi-structured open-ended interviews. CPs received a $25 gift card for each assessment. The quantitative data collection included demographic information and self-administered close-ended questionnaires.

The overall mixed methods approach is described in Figure 1. First, a qualitative thematic analysis combined with matrix analysis (60) was undertaken to develop cross-cutting patterns from interviews with CPs. Three coders independently read interview transcripts in “open coding,” where each analyst inductively identified relevant excerpts with a provisional label (61). Analytic memos were written regularly during open coding to connect emergent content related to community reintegration. Case summaries were compared using a data matrix to further identify themes and achieve data saturation. A codebook was developed and refined until a shared understanding was achieved. Next, at least two team members independently coded each transcript; pairs met in person to review the double-coded transcripts and resolve discrepancies through consensus. Qualitative data was coded and analyzed using NVivo (62). Subsequently, themes were aligned with concepts from the caregiver burden model (i.e., self-perception, multi-faceted strain, disclosure, navigation, resources, needs, and strategies).

Consistent with the study aims, cases were sorted using baseline ZBI scores. Cases were categorized into high burden (ZBI score ≥ 13, 10/34, 29.4%) and low burden (ZBI score <13, 24/34, 70.6%) groups for focused analysis of interview transcripts according to their baseline burden score. Combining 12-month scores for burden with baseline burden scores, cases were categorized as staying high (4/27, 14.8%), low (17/27, 63.0%), shifting from high to low burden (4/27, 14.8%), or low to high burden (2/27, 7.4%). Seven CPs did not complete surveys at 12 months.

For the quantitative measures, descriptive statistics (e.g., mean, frequencies) were calculated to characterize the sample in terms of demographic variables and psychosocial outcomes. To examine associations between CP characteristics and outcomes, Pearson correlations were calculated among PSS, SFI, SWLS, Zarit Burden, and Quality of Relationship Scale scores. Additionally, to compare CPs reporting high burden with those reporting low burden, independent t-tests were used to compare these groups on PSS, SFI, SWLS, and Quality of Relationship Scale scores. To account for multiple comparisons increasing the possibility of a Type I error, a False Discovery rate adjustment was made (63).

In the next stage of analysis, baseline ZBI scores were used to categorize cases into high-burden (ZBI score ≥ 13, 10/34, 29.4%) and low-burden (ZBI score <13, 24/34, 70.6%) groups. Combining 12-month scores for burden with baseline burden scores, cases were categorized as staying high (4/27, 14.8%), low (17/27, 63.0%), shifting from high to low burden (4/27, 14.8%), or low to high burden (2/27, 7.4%). Seven CPs did not complete surveys at 12 months. In the last phase of analysis, we selected exemplar cases that illustrated cross-cutting findings and changes in burden over time.

Of the 36 CPs, a majority were women (72.2%) and married or partners (72.2%) to the Veteran for whom they provided care (see Table 1). Consequently, most CPs lived with the Veteran (63.9%) and, on average, had known the Veteran for a considerable length of time, mean = 15.1 years (SD = 10.0, range = 0.8, 34). Only 11.1% of CPs had been caregiving for the Veteran for less than a year, while 52.8% had been a caregiver for more than 5 years. Also, 58.3% of CPs were parents or guardians to a child under the age of 18 years. Over a third (36.%) reported military service. One respondent reported participation as a paid caregiver in the VHA PCAFC.

As expected, burden was significantly positively correlated to perceived stress (r = 0.50, p = 0.003). Additionally, burden was significantly inversely related to secure flourishing (r = −0.60, p < 0.001) and relationship quality (r = −0.45, p = 0.008). Though an inverse trend between burden and life satisfaction was observed, this correlation was non-significant (r = −0.26, p = 0.14).

Descriptive statistics are shown in Table 2 for survey measures completed by CPs. As expected, CPs with high burden had significantly higher stress (p = 0.02), lower secure flourishing (p = 0.02), and marginally significantly lower life satisfaction (p = 0.08) and relationship quality (p = 0.06).

As shown in Table 3, CPs did not access many of the resources available. The most frequently accessed resources were a religious or spiritual network and informal information sources, such as websites and articles. In contrast, formal, structured programs and resources, such as loans, support groups, case managers, caregiving training, and stipends, were rarely accessed.

Based on the themes and case descriptions that follow, Figure 2 depicts how burden was conceptualized in this study. The diagram retains the dynamic relationship between self-perception, multi-faceted strain, and change over time but adds factors that increase or decrease the burdens that are specific to CPs to Veterans.

CPs expressed a wide range of feelings and perceptions associated with their caregiving role. Negative emotions named included stress, anger, overwhelmed, frustrated, tired, or drained. For example, P4032 stated, “I feel like I run around like a chicken with my head cut off. So as a spouse, I feel like it's very overwhelming;” P4004 explained, “So yeah, I get stressed out. Do I know everything about what to do about everything all of the time? Absolutely not.”

However, CPs described their own qualities that were helpful in managing their caregiver role. These qualities included being adaptable, understanding the military experience, accepting their situation, being solution-focused, staying calm, and being a positive, caring person. As an example of adaptability being an asset, P4021 described, “I'm used to being flexible… it was always important to me for my husband to be happy with whatever job he's working for after the military, so if that meant we need to move a certain city, that's okay. I work in healthcare, so it's not as hard for me to find jobs.” Similarly, another CP relied on her acceptance and adaptability for managing the caregiver role. P4017 stated, “It's not a pity thing… When you sign on and you know that this is a lifestyle, you are choosing, and the choice was made that our marriage and we know that like his Marine Corps job is what's going to drive. And we were in that together. So but as a result, that means that I am not going to have a career… There is a lot of fluidity, and you have to be willing to just go with the flow and let his stuff drive it and not yours, if that makes sense.” P4006 described herself as a “fixer” meaning she will “try to figure out well, how can we fix this? Well, how can we make it better?” She viewed this as a strength in her role as a caregiver. Similarly, P4038 described herself as “a person that loves to take care of people” and “the bubbly, cheery, happy-go-lucky person. Day in and day out, will find the positive in everything.”

As noted by Liu et al. (30), CPs manage caregiving responsibilities alongside their other social roles. CPs reported that they helped their Veteran with daily medications, medical appointments, physical safety (i.e., falling), emotional support, and meals. One CP described monitoring her husband's mood and providing support when he appeared to need it. P4006 stated, “So at this point, that's what I really feel like that my role is really just to make sure that he has a loving support at home and that he knows that he's accepted no matter what and… recognize that oh okay, something is a little off… Either he's not communicating, or he seems a little withdrawn and then just trying to gently talk to him about are you just tired.” Similarly, when asked “Are you like worried about him?” P4012 responded, “It's kind, it's almost like a motherly worried. Because I can tell that he's anxious and that he's feeling that anxiety.” P4011 described her caregiving stress in the following way: “It can be a very, very tiring struggle sometimes to help everybody and make sure that he is still okay and still on the schedule that he needs or making it to the appointment or remembering to eat.” In addition to monitoring emotional changes, CPs reported that Veterans had angry outbursts that were hard to manage. P4020 explained: “I'm trying to keep him calm. You know… It's upsetting because he'll get angry and be upset more, which can cause the kids to get upset. Yeah, it's stressful.” Some of these CPs primarily managed household finances, daily household decisions, and household moves, or were the only driver in the household. For example, P4003 stated “So I take care of all of the financial stuff at our house.” Similarly, P4011 explained, “And I'm like the stick in the middle that controls everything (laughter). Yes. Very much so. I kind of rule everything, and I didn't like to take that spot, but I do pretty good at it (laughter).”

CPs played a crucial role in supporting help-seeking behavior, including encouraging Veterans to disclose information with appropriate timing and context. P4016 explained how her husband “was so closed off” that he did not respond to initial mental health outreach, but that her presence at therapy appointments was beneficial: “I was able to kind of explain certain situations… that he maybe he didn't even realize was happening.”

CPs describe how they helped Veterans navigate through challenges related to accessing health services, scheduling appointments, and paying for health services. One CP expressed how it “was easier to navigate with two brains than one” (P4013). This CP emphasized caring for his spouse: “No, it's not a burden. I'm not even stressed either, because of the way I deal with stress, which is another thing that she is mad about, how I'm different from her on that aspect.”

P4025, a CP with high self-reported stress, described in detail attempts to help:

This CP explained the importance of persistent monitoring coupled with empathy about the Veteran's anxiety about attending scheduled appointments. Other CPs discussed how they witnessed “getting the runaround” (P4011) when trying to access services; P4004 explained that it was necessary to reach out to the Patient Advocate for assistance.

In terms of resources that would aid CPs, respondents discussed a range of needs and strategies that they used. Several CPs suggested that focused courses or workshops delivered after military separation occurs would be beneficial with specific courses dedicated to CPs supporting Veterans. Topics that were suggested included communication skills, awareness of mental health issues, and support in navigating the VA healthcare system. P4020 suggested a coupled-focused course: “No, nobody ever contacted me and said this is what PTSD looks like—like I think that it should be mandatory for couples to go through classes before deployments and after deployments, and you know.” Other CPs expressed the need for consistency when it comes to health care services and courses to understand VA benefits, including disability and insurance from a CP perspective. P4002 proposed presentations about supporting a Veteran be held at local community centers to aid in caring efforts.

Table 3 summarizes which types of resources CPs found useful. The most accessed resources, structured wellness activities, informal information sources, and participation in spiritual or religious networks, were reported to be helpful.

The following offer exemplar cases that illustrate high and low burdens. To demonstrate within-case changes, we likewise offer cases that remained as either high or low burden or changed their burden-level at 12 months.