Effect of cohabitating partners on the development of posttraumatic stress symptoms after emergency department visits for stroke and transient ischemic attack

Abstract

Objective:

Partners can be beneficial for patients experiencing stressful health events such as a stroke/TIA. During such events, however, partners may exacerbate early distress. The present study tested whether having a cohabiting partner modified the association between patients’ early perceptions of threat (e.g., feeling vulnerable, helpless) and longer-term posttraumatic stress symptoms (PTSS).

Methods:

Participants (N = 328) were drawn from an observational cohort study of patients evaluated for stroke/TIA at an urban academic hospital between 2016–2019. Participants self-reported emergency department (ED) threat perceptions and PTSS secondary to the stroke/transient ischemic attack at three days and one-month post-event.

Results:

Cohabiting partner status modified the association of ED threat with early PTSS. Patients with a cohabiting partner exhibited a positive association between ED threat and early PTSS, B = 0.12, p < .001; those without a cohabiting partner did not, B = 0.04, p = .067. A cohabiting partner was protective only for patients who initially reported low levels of ED threat, as patients with a cohabiting partner who reported low levels of ED threat also had lower early PTSS, B = −0.15, p = .016; at high levels of ED threat, a cohabiting partner was not protective, B = −0.02, 95% CI [−0.14, 0.09], p = .68. ED threat was associated with PTSS at one month, B = 0.42, p < .001, but cohabiting partner status did not modify the association.

Conclusions:

ED threat perceptions were positively associated with early PTSS only for patients with a cohabitating partner. For patients who do not initially experience a stroke/TIA event as threatening, cohabiting partners may help patients maintain psychological equanimity.

Introduction

Having a romantic relationship is one of the most consistent psychosocial predictors of better mental and physical health (J. K. Kiecolt-Glaser & Newton, 2001; Janice K Kiecolt-Glaser & Wilson, 2017; Robles et al., 2014). In particular, partners are considered a primary source of social support for patients experiencing serious health events (United Hospital Fund, 2012) such as a stroke or transient ischemic attack (TIA). For example, empathic partner support has been associated with beneficial outcomes such as improvements in physical functioning in patients with osteoarthritis (Hemphill et al., 2016). We have shown, however, that there are circumstances under which partners can instead worsen outcomes by exacerbating patients’ perceptions of threat from a medical event (e.g., feeling vulnerable or helpless), such as when they are present in the highly stressful emergency department (ED) during evaluation and/or treatment for an acute medical event (Cornelius et al., 2019a; Cornelius et al., 2019b; Cornelius et al., 2020). This early amplification of patients’ threat perceptions secondary to a medical event may have serious adverse long-term consequences, as threat perceptions in the ED are associated with increased risk for long-term distress—in particular, posttraumatic stress symptoms (PTSS; Cornelius et al., 2018; Meli et al., 2018)—which, in turn, predict elevated risk for medical event recurrence and mortality (Edmondson et al., 2012). Despite the link between partner presence in the ED and patient threat perceptions, it remains possible that living with a partner buffers patients from developing clinically significant psychological distress in the long-term (Borstelmann et al., 2015). Conversely, partners – especially those living with the patient – may exacerbate the link between early threat and long-term distress in patients due to dyadic processes such as co-rumination or stress contagion (Buchanan et al., 2012; Cornelius; Engert et al., 2014; Gump & Kulik, 1997). No research has examined whether cohabitating partners dampen or exacerbate the association between patients’ ED threat perceptions and PTSS after stressful medical events. We therefore sought to test two competing hypotheses: (1) living with a partner weakens the link between patients’ early threat perceptions in the ED and PTSS, versus (2) living with a partner strengthens the link between patients’ early threat perceptions in the ED and PTSS. We examined these possibilities for both early PTSS (acute stress symptoms approximately three days post-event) and PTSS at least one month later (the symptom duration required for PTSD diagnosis).

Psychological distress is common in patients during and after an acute cardiovascular event. In the ED, approximately 30% of patients evaluated for an acute cardiac event report significant feelings of vulnerability (White et al., 2017) and approximately 10% worry that they will die. As many as 23% of patients develop clinically significant PTSS in the year following stroke/TIA (Edmondson et al., 2013). Clinically significant PTSS secondary to an acute medical event are associated with poor adherence to medications for managing secondary risk (Kronish et al., 2020; Kronish et al., 2012) and a doubling of risk for event recurrence and mortality (Edmondson et al., 2012). Due in part to this link between psychological distress and long-term health outcomes, a growing body of research has sought to understand the etiology of medical event-induced PTSS (Vilchinsky et al., 2017) and to develop effective preventive methods (Birk et al., 2019).

Greater perception of threat due to a medical event during ED evaluation (e.g., feeling vulnerable or helpless) is a risk factor for developing PTSS following an acute medical event (Cornelius et al., 2018; Meli et al., 2019; Meli et al., 2018). Because romantic partnerships are strongly associated with lower levels of psychological distress, both cross-sectionally and prospectively (Braithwaite & Holt-Lunstad, 2017; Karney & Bradbury, 1995; Proulx et al., 2007; Whisman & Baucom, 2012), it is often recommended to have someone present to provide social support during acutely stressful health situations (e.g., ED visits; United Hospital Fund, 2012). Indeed, positive forms of social support can buffer against the development of PTSS in other types of traumatic events (e.g., motor vehicle accidents; Robinaugh et al., 2011). Nevertheless, we have shown that bringing a close other, such as a partner, to the ED, may not be universally helpful. Patients who are in the ED with close others (e.g., a partner), feel more threatened by the medical event than those who arrive with non-close others (e.g., a neighbor), or even those who arrive alone. This, in turn, may be associated with medically induced PTSS at least one month – or more – later (Cornelius et al., 2019b).

The partners of patients are also greatly impacted by the ED experience and the acute medical event that brought them there (Dalteg et al., 2011; de Miranda et al., 2011; Fait et al., 2017; Tulloch et al., 2020; Vilchinsky, 2017; Zivin & Christakis, 2007). We have shown that close others are more likely to require emotional support from patients, and to increase patient anxiety, than non-close others (Cornelius et al., 2019a). The mechanisms of the early impact of partner presence on patient distress (Cornelius et al., 2019b) may be due to processes such as stress contagion (Cornelius; Gump & Kulik, 1997), as partners may be highly distressed during ED evaluation. Indeed, there is evidence that one person’s physiological stress response (e.g., cortisol increases following an acute stressor) may influence that of another person, and this stress-contagion effect is stronger for in-person than virtual dyads and stronger among intimate partners than strangers (Buchanan et al., 2012; Engert et al., 2014).

Questions regarding long-term effects of partner presence (i.e., cohabitation) on the threat-distress link remain unanswered. Might partners be beneficial for distress in the long term? That is, even if partners might increase patients’ threat perceptions during the potentially traumatic event in the ED, can they mitigate the longer term consequences of threat for the development of PTSS in the days and weeks following? Alternatively, might partners unwittingly exacerbate the effect of early threat perceptions on subsequent PTSS?

Given these competing possibilities, the present study sought to address this research gap in a sample of patients evaluated for stroke/TIA in the ED. Specifically, we sought to answer the following question: Does having a cohabiting partner modify the association between patients’ threat perceptions in the ED and subsequent distress, including early PTSS (acute stress symptoms several days post-event (Bryant et al., 2000) and PTSS occurring at least one month later? If so, do partners weaken or exacerbate this ED threat—PTSS association?

Methods

Participants

Participants were drawn from a prospective cohort study examining the development of PTSS secondary to a stroke/TIA. Participants were patients presenting to the NewYork Presbyterian Hospital (NYPH) ED for evaluation for a stroke/TIA between 2016 and 2019. Eligible patients were at least 18 years of age, were evaluated for stroke/TIA in the NYPH ED as determined by an admitting neurologist and were fluent in English or Spanish. Patients with terminal non-cardiovascular illness (life expectancy of less than one year), severe mental illness requiring urgent psychiatric hospitalization or intervention, alcohol or substance abuse that would prevent the ability to complete the study protocol, severe stroke (NIH Stroke Scale [NIHSS] >14 and/or significant speech or cognitive impairment), or lack of availability for follow up were excluded from this study. The Columbia University Irving Medical Center Institutional Review Board approved all study procedures, and all participants provided written informed consent before completing the study.

Procedure

Research assistants (RAs) contacted the clinical staff on the NYPH stroke service to determine preliminary eligibility for the parent study. Eligibility included patients being evaluated for stroke/TIA, who were capable of completing study protocols (no cognitive impairments or history of substance abuse). After obtaining clinician permission, RAs approached patients to offer participation in the study while they were in the ED, in a hospital bed, or at the Rapid Access Vascular Evaluation-Neurology (RAVEN) clinic after their initial ED visit. After providing written informed consent, participants completed the baseline self-assessment interview. If enrolled in the ED or if the baseline was unfinished, the baseline interview was completed either by telephone (if discharged) or in-person while in a hospital bed (if admitted from the ED) a median of three days after enrollment. The medical record number was recorded to track enrolled participants and for medical chart extraction purposes. A telephone interview was administered at a median of 53 days following completion of the baseline interview (note that this window begins at one-month post-discharge, hence we call it the one-month interview). Medical covariates (e.g., stroke severity) were extracted from patient medical records.

Measures

Early posttraumatic stress symptoms.

Early PTSS were self-reported a median of three days following ED enrollment (i.e., prior to the one-month elapsed time necessary for determining PTSD diagnosis) using the 19-item Acute Stress Disorder Scale (ASDS; Bryant et al., 2000). This scale measures early symptoms associated with risk for developing subsequent PTSS and has excellent properties, including strong internal consistency and test-retest reliability (Bryant et al., 2000). Response options range from 1, “Not at all,” to 5, “Very much,” such that higher scores indicate greater levels of acute stress symptoms, Cronbach’s α = 0.85.

Posttraumatic stress symptoms.

PTSS due to the stroke/TIA were assessed at a minimum of one month later using the PTSD Checklist for DSM-5 (PCL-5). This 20-item scale is a reliable and valid tool for assessing PTSS in clinical and research settings (Weathers et al., 2013) and has been validated for use in cardiovascular patients (Sumner et al., 2015). The PCL-5 assesses the presence, frequency, and severity of PTSS. Response options range from 0, “Not at all,” to 4, “Extremely,” such that higher scores indicate greater levels of PTSS, Cronbach’s α = 0.92.

Cohabiting partner status.

This variable was dummy coded, such that participants who self-reported having a partner or spouse and living with this person were coded as 1, “Yes,” or 0, “No.”

ED threat perceptions.

We assessed threat perceptions (e.g., helplessness, fear, vulnerability) due to hospitalization for stroke/TIA using the seven-item ED Threat Perceptions questionnaire (Cornelius et al., 2018) at the inpatient interview, a median of three days following ED enrollment. This scale assesses perceived threat due to a medical event while a patient is in the ED or shortly thereafter, has been validated for use in both English- and Spanish-speaking patient populations. All items are available in the validation paper (Cornelius et al., 2018). Response options range from 1, “Not at all,” to 4, “Extremely,” such that higher scores indicate greater levels of threat, Cronbach’s α = 0.87.

Covariates.

Covariates were selected a priori and for consistency with the larger parent cohort analyses. These included baseline assessments of self-reported gender, age, race/ethnicity, partner status, pre-existing PTSS(Weathers et al., 1993) (only individuals who self-reported a prior trauma, via the Life Events Checklist [Gray et al., 2004], had non-zero symptoms scores), and pre-existing depression (Kroenke et al., 2009). Stroke severity was assessed by patients’ medical providers using the 11-item NIHSS, a clinical tool used to evaluate neurological status (Richardson et al., 2006).

Data Analysis Strategy

We conducted linear regression analyses in SAS v. 9.4 predicting (1) early PTSS (acute stress symptoms) and (2) long-term PTSS (log-transformed prior to analysis due to non-normal distributions). For each outcome, we estimated two models. The first included the above specified covariates (i.e., gender, age, race/ethnicity, partner status, pre-existing PTSS and depression, stroke severity) and the main effects for cohabiting partner status and ED threat perceptions, and the second also included the multiplicative interaction of these two terms. For each outcome, if a significant interaction was found, this result was followed up with a simple-slopes analysis examining the association of ED threat perceptions with the relevant outcome variable for patients with a cohabiting partner and for those without.

Results

Of the 515 participants in the parent cohort, 328 (63.69%) provided complete data for the present analysis. Those participants who were missing data on predictor variables, outcome variables, or covariates had significantly greater early PTSS (Median = 32.00 versus 27.50, p < .001) and PTSS at one month (Median = 8.00 versus 5.00, p = .001), higher levels of pre-existing PTSS (Median = 5.50 versus 2.00, p = .017) and pre-existing depression (Median = 6.00 versus 3.00, p < .001), and greater stroke severity (Median = 3.00 versus 2.00, p = .014). Female participants, white participants, and Hispanic participants were less likely to be retained at follow-up, relative to male participants and participants of non-White race and non-Hispanic ethnicity, respectively, ps < .001. ED threat perceptions, age, gender, and partner status and cohabitation did not differ between included and excluded participants.

Mean age was 61.43 (SD = 15.43), 44.82% lived with a partner, and 47.87% were female. More than one third of participants (38.72%) were Hispanic, and 25.91% were Black. Demographic data are available in Table 1, and correlations between distress variables are in Table 2. Full model results are detailed in Table 3. ED threat perceptions were not associated with cohabiting partner status, t(326) = 0.48, p = .63 (with a cohabiting partner: M = 1.23, SD = 0.81; without a cohabiting partner: M = 1.19, SD = 0.80).

Table 1.

Demographic data.

		Mean (SD), N (%), or Median [IQR]
Age		61.43 (15.43)
Sex	Male	171 (52.13%)
	Female	157 (47.87%)
Race/Ethnicity	White	84 (25.61%)
	Black	85 (25.91%)
	Hispanic	127 (38.72%)
	Other	32 (9.76%)
Partner		167 (50.91%)
Cohabiting partner		147 (44.82%)
NIHSSa		2.00 [1.00, 4.00]
Baseline depression symptoms		3.00 [0.01, 7.00]
Pre-existing PTSSb		2.00 [0.00, 13.00]
ED threat perceptions		1.14 [0.57, 1.86]
Early PTSSc		27.50 [22.00, 36.00]
PTSS at one month		5.00 [1.00, 14.00]

^aNIH Stroke Scale

^bPosttraumatic stress symptoms

^cAcute stress symptoms

Table 2.

Correlations between distress variables.

	Baseline depression symptoms	Pre-existing PTSS	ED threat perceptions	Early PTSS	PTSS at one month
Baseline depression symptoms	1.00
Pre-existing PTSSa	0.45**	1.00
ED Threat Perceptions	0.26**	0.23**	1.00
Early PTSSb	0.59**	0.57**	0.39**	1.00
PTSS at one month	0.43**	0.45**	0.37**	0.63**	1.00

^*p < .01,

^**p < .001.

^aPosttraumatic stress symptoms

^bAcute stress symptoms

Table 3.

Full results for the association of ED threat perceptions, cohabiting partner status, and their interactions with early PTSS and PTSS at one-month, natural log-transformed.

		Early PTSSd	PTSS at one month
		B (se)	B (se)
Model 1
Agea		−0.01 (0.01)	−0.13 (0.06)+
Sex	Male	Ref	Ref
	Female	0.00 (0.03)	0.18 (0.19)
Race/Ethnicity	White	Ref	Ref
	Black	0.05 (0.04)	0.38 (0.25)
	Hispanic	0.07 (0.03)+	−0.01 (0.23)
	Other	0.19 (005)**	0.91 (0.34)*
NIHSSa,b		0.08 (0.04)+	0.64 (0.28)+
Baseline depression symptoms		0.22 (0.03)**	0.51 (0.19)*
Pre-existing PTSSa,c		0.08 (0.01)**	0.30 (0.08)**
Partner		0.13 (0.05)*	0.44 (0.39)
Cohabiting partner		−0.08 (0.06)	−0.33 (0.39)
ED threat perceptions		0.08 (0.02)**	0.42 (0.12)**
Model 2
Cohabiting partner * ED threat perceptions		0.08 (0.03)*	0.26 (0.22)

⁺p < .05,

^*p < .01,

^**p < .001

^aDivided by 10

^bNIH Stroke Scale

^cPosttraumatic stress symptoms

^dAcute stress symptoms

Early PTSS

In the first model, as hypothesized, ED threat perceptions were positively associated with early PTSS, B = 0.08, 95% CI [0.04, 0.11], p < .001; cohabiting with a partner was not, B = −0.08, 95% CI [−0.19, 0.03], p = .14. In the second model, these main effects were qualified by a significant interaction between ED threat and cohabiting with a partner, B = 0.08, 95% CI [0.01, 0.14], p = .032. This pattern indicated that patients cohabiting with a partner demonstrated a significant and positive association between ED threat perceptions and early PTSS, B = 0.12, 95% CI [0.07, 0.17], p < .001, but patients without a cohabiting partner did not show this association, B = 0.04, 95% CI [−0.00, 0.08], p = .067.

An alternative way of viewing the interaction between ED threat and cohabiting with a partner is to focus primarily on the association of cohabitation with PTSS—rather than on how cohabitation may alter the expected course of PTSS development (i.e., the threat-PTSS link). A cohabiting partner (versus no cohabiting partner) was not associated with early PTSS at high levels of perceived threat, B = −0.02, 95% CI [−0.14, 0.09], p = .68, but was significantly associated with lower early PTSS at low levels of perceived threat, B = −0.15, 95% CI [−0.27, −0.03], p = .016 (see Figure 1).

Fig. 1.

Association of emergency department (ED) threat perceptions with early PTSS for individuals with versus without a cohabiting partner.

PTSS One Month Later

In the first model, as hypothesized, ED threat perceptions were positively associated with PTSS occurring at least one month later, B = 0.42, 95% CI [0.18, 0.65], p < .001; cohabiting with a partner was not, B = −0.33, 95% CI [−1.10, 0.44], p = .40. The interaction between ED threat perceptions and cohabiting with a partner was not significant, B = 0.26, 95% CI [−0.18, 0.70], p = .25.

Post-hoc Analysis

To better understand long-term implications of cohabiting partners for the development of stroke/TIA-induced PTSS from early threat perceptions in the ED, we conducted a mediation test in which ED threat perception score, cohabiting partner status, and their interaction were the predictors, PTSS at one month was the outcome, and early PTSS was the hypothesized mediator. In this model, the association between early PTSS and PTSS at one month was positive and significant, B = 2.57, 95% CI [1.83, 3.31], p < .001. For patients with a cohabiting partner, the indirect association of ED threat perceptions with PTSS at one month, via early PTSS, was thus a×b = 0.12×2.57 = 0.31. Using joint significance testing, we can conclude that this is a significant indirect effect (MacKinnon et al., 2002). For those participants without a cohabiting partner, this association was a×b = 0.04×2.57 = 0.10.

For participants with high levels of ED threat perceptions (+1 SD), the indirect association of a cohabiting partner with PTSS at one month, mediated via early PTSS, was a×b = −0.02×2.57 = −0.05. For those participants reporting low levels of ED threat perceptions (−1 SD), this association was a×b = −0.15×2.57 = −0.39, and was statistically significant (MacKinnon et al., 2002).

Discussion

The present study tested whether, and how, having a cohabiting partner modified the association between threat perceptions in the ED and subsequent development of PTSS in a sample of patients evaluated for stroke/TIA. We posited that partners may buffer this association (i.e., reduce or prevent development of PTSS), or, conversely, might exacerbate this link due to dyadic processes as they unfold in the days and months after the neurological event. Results supported the latter possibility. Specifically, relative to patients without cohabitating partners, patients with cohabiting partners showed a significantly stronger association between ED threat perceptions and early PTSS in the first several days after the ED visit (i.e., symptoms indicative of acute stress; see Figure 1). This finding supports the idea that partners are not always beneficial for patients’ mental health in the early period after the acute medical event – rather, the stronger association between ED threat perceptions and early PTSS in these patients effectively eliminated any benefit of having a cohabiting partner (i.e., although cohabiting partners were not helpful for these patients, they were not harmful either). Yet, when ED threat perceptions were low, patients with cohabiting partners showed significantly lower early PTSS relative to patients without cohabitating partners. This finding suggests that partners may be helpful for patients’ mental health provided that the patients’ initial perception of threat during the potentially traumatic event was low (and, consequently, they were not at high risk of developing PTSS). The finding that cohabiting partners do not yield benefit for patients with high initial distress, and may cement early distress over time, is both novel and important.

Some partners may buffer the development of longer-term distress by preventing early PTSS. This is in line with research evidencing an association between partner status and mental wellbeing (Braithwaite & Holt-Lunstad, 2017; Karney & Bradbury, 1995; J. K. Kiecolt-Glaser & Newton, 2001; Janice K Kiecolt-Glaser & Wilson, 2017; Proulx et al., 2007; Robles et al., 2014; Whisman & Baucom, 2012) and that social support can buffer the development of PTSS following traumatic events (Robinaugh et al., 2011). This was true, however, only for patients who perceived less threat during ED evaluation (see Figure 1). It is possible that partners, who are also greatly impacted by a patient’s stroke/TIA (Draper & Brocklehurst, 2007; Green & King, 2007, 2009), are better equipped to provide helpful forms of support if the patient perceives only a minimal-to-moderate level of threat during the acute medical event. Calmer patients may lead their partners to remain calm, and as a result, even less typically helpful forms of support provided by partners may be useful for these relatively unthreatened patients. Stated otherwise, patients who do not perceive the experience of the acute medical event and ED environment as particularly threatening may be able to reap the benefits of less skilled support from their partners. Furthermore, patients who are less distressed may also require less support (Cornelius et al., 2019a; Rafaeli & Gleason, 2009).

The question is thus, why might this protective factor break down when patients perceive the ED and acute medical event as especially threatening? If patients perceive threat, unskilled or anxiety-provoking support provided by distressed partners may be harmful (Cornelius et al., 2019a; Rafaeli & Gleason, 2009). Indeed, partners are also highly distressed following patients’ medical events (Fait et al., 2017; Vilchinsky, 2017), and social psychological theories of stress contagion, co-rumination, and more suggest that an early distress response in one couple member can elicit or exacerbate adverse outcomes in the other (Buchanan et al., 2012; Cornelius, 2021; Engert et al., 2014; Gump & Kulik, 1997; Müller et al., 2019). Critically, the present findings suggest that this detrimental process may continue through the sensitive early period in the days after the acute medical event, such that the patient remains at increased risk for the subsequent development of PTSS (Meli et al., 2019) and is prevented from reaping the benefit of having a cohabiting partner. Our observation that patients who had cohabitating partners showed a more robust link between their ED threat perceptions and their early PTSS (acute stress symptoms) is consistent with this possibility.

It is perhaps unfortunate that only patients who do not perceive a medical event as threatening evidence this potential long-term psychological benefit from a cohabiting partner; yet, this is also an opportunity to design couple-targeted interventions that can harness the power of cohabiting partners to improve patients’ wellbeing (Martire & Helgeson, 2017; Martire et al., 2010). Future research is needed to explore whether interventions that target both the patient and the partner in the ED and early post-stroke/TIA are beneficial in alleviating stress and anxiety experienced from the health event. By offering formal, skilled support in the ED to couples, there may be potential to directly mitigate distress and other adverse psychological effects in patients and partners. Patients may also indirectly benefit when partners experience higher levels of psychological well-being (e.g., if these partners are better equipped to support patients upon returning home). Skilled support may reduce the harm of initial threat perceptions by providing a stabilizing resource to guide both patient and partner from the acute ED phase of medical care into the following phase of care as the patient returns home from the hospital.

Limitations

Data were drawn from a sample of stroke/TIA patients at an urban academic ED, and findings may not generalize to other patient populations (e.g., rural or suburban areas, other medical conditions). Our prior work showing links between close others and ED threat perceptions have examined partner presence with the patient while in the ED (Cornelius et al., 2019a; Cornelius et al., 2019b). Unfortunately, these data were not available in the present study. Future studies should examine mechanisms underlying protective and harmful dyadic processes (e.g., stress contagion, social support; Cornelius, 2021; Gump & Kulik, 1997; Robinaugh et al., 2011) that are hypothesized to contribute to the development or maintenance of patient distress by collecting data over time in both patients and partners (e.g., ecological momentary assessment). As this study was not designed to examine patients’ relationships, no additional variables were available to assess patient and partner relationships (e.g., satisfaction, relationship length). The present analyses may have been limited by the set of planned covariates that could influence the development of PTSS (e.g., stroke severity but not number of comorbidities); however, number of medical comorbidities has previously been shown to be unrelated to the development of PTSS after acute medical events (e.g., myocardial infarction). Finally, this study is observational and there may be unmeasured couple-level variables (e.g., poor dyadic coping, both early post-stroke/TIA and long-term) that contributed to these correlational associations.

Conclusion

Prior literature is inconclusive about whether partners are generally protective or harmful concerning patients’ risk of developing psychological distress after acute medical events (Borstelmann et al., 2015; Cornelius et al., 2019a; Cornelius et al., 2019b). We sought to address this question by elucidating whether living with a partner versus not living with a partner is associated with a stronger or weaker link between initial threat perceptions during the ED visit for stroke/TIA, a potentially traumatic event, and PTSS in the days and weeks after the event. On the one hand, there was evidence consistent with the possibility that cohabitating partners may exacerbate the established risk associated with ED threat perceptions for the development of PTSS (Meli et al., 2019), effectively eliminating any benefit of having a cohabiting partner for long-term distress. There was also evidence consistent with the notion that partners were helpful; they may protect against the development of PTSS over time, provided that the patient does not exhibit elevated ED threat perceptions. The study lends further support and direction to intervention research that targets the psychological well-being of couples in the ED. Future research should examine couple-level variables that impact cohabiting partners and patients, including shared threat perceptions, social support, dyadic coping, and more, and should work to uncover individual and dyadic mechanisms by which partners impact patient outcomes in the early days, weeks, and months following a stressful health event.

Highlights.

• Patients in the ED for stroke/TIA can develop posttraumatic stress symptoms (PTSS).

• Cohabiting partners of patients with low distress in the ED may be beneficial.

• Cohabiting partners may exacerbate ED distress, contributing to long-term PTSS.