Person-centered analysis of psychological traits to explain heterogeneity in patient-reported outcomes of coronary artery disease– the THORESCI study

Highlights

• Four groups were found: Low distress, Passive coping, Active coping, High distress.

• Patient-reported outcomes were significantly different between the subgroups.

• A person-centered approach is useful in explaining heterogeneity in PCI recovery.

Abstract

Background

Heterogeneity in the prognosis of coronary artery disease (CAD) patients may be explained by relatively stable individual psychological differences. Therefore, we studied multiple personality and coping traits using a person-centered approach, and examined the predictive value of this approach for patient-reported outcomes.

Method

657 CAD patients (age = 66.39 ± 10.6; 79% men) completed multiple self-report questionnaires focusing on demographics, negative affectivity and social inhibition (DS14), neuroticism and extraversion (EPQ), resilience (DRS-15), and coping styles (CISS) after undergoing percutaneous coronary intervention. Depressive symptoms (PHQ-9), anxiety (GAD-7), and treatment adherence (MOS) were assessed at 6 months follow-up. Clinical information was extracted from patients’ medical records.

Results

A step-3 latent class analysis identified four subgroup profiles: Low distress (31%), Passive coping (21%), Active coping (20%), and High distress (28%). For all patient-reported outcomes, overall significant differences between the subgroups were observed (p-values < .05). The High distress profile was associated with the highest levels of emotional distress (d’s > .94), and lowest levels of positive mood (d = −1.02) and treatment adherence (d = −2.75) at follow-up. Patients with an Active coping profile also experienced increased emotional distress (d’s > .50), but participated in cardiac rehabilitation most often (d = .13), and reported high levels of positive mood (d = −1.02). Patients with a Passive coping profile displayed few emotional problems after six months (d’s < .30), but participation to cardiac rehabilitation was relatively low (d = .04).

Conclusions

This study revealed four distinct psychological latent subgroups, which were predictive of patient-reported outcomes. The results indicate that a person-centered approach is useful in explaining heterogeneity in recovery from PCI, and may enhance personalized medicine in patients with CAD.

Introduction

Psychosocial factors have been associated with the morbidity and mortality of coronary artery disease (CAD) (Lichtman et al., 2014, Rozanski et al., 2005, Yusuf et al., 2004). Several recent meta-analyses have indicated that negative emotions, such as depression (Nicholson et al., 2006) and anxiety (Roest et al., 2010), but also more stable traits, such as Type D personality (combination of negative affectivity and social inhibition) (Beutel et al., 2012, von Kanel, 2012), are associated with adverse events and mortality. In addition, chronic stressors, such as a lower socio-economic status (Vathesatogkit et al., 2014) and a lack of social support (Barth et al., 2010) may worsen CAD. However, these meta-analyses also reported substantial heterogeneity in when and how psychosocial factors affect the prognosis of CAD. While several studies in CAD patients showed an association between psychosocial factors, such as anxiety and depression, with major adverse cardiovascular events during follow-up (de Jager et al., 2018, Meyer et al., 2015), other studies did not find these associations (Pelletier et al., 2015). Moreover, a recent study showed large heterogeneity in the prognostic value of individual depressive symptoms in patients with CAD across sex and age subgroups (de Miranda Azevedo et al., 2018).

This heterogeneity may have several causes, including age differences, comorbid medical conditions and other differences in study characteristics (Denollet, 2013). The risk of adverse events is already higher in older patients due to an aging heart and age-associated conditions, such as kidney disease and anemia (Alexander et al., 2007, Shih et al., 2011, Tay et al., 2008). Furthermore, the choice of endpoint is a crucial determinant of the prognostic effect of a risk factor (Mann and Felker, 2014). For example, a recent study showed that Type D personality may be more related to fatal and non-fatal cardiac events than to all-cause mortality, and has different effects in different age-groups (Kupper and Denollet, 2016).

Heterogeneity in clinical outcomes may also be explained by relatively stable individual differences in personality (Chapman et al., 2011). Specific sets of interrelated thoughts, feelings, and behaviors might affect the propensity to experience lower or higher levels of psychological distress (Chapman et al., 2011). Patients scoring high on neuroticism are more vulnerable to experience negative emotions such as anxiety, depression, and anger after a cardiac event (Terracciano et al., 2008). In contrast, resilient patients seem to report better emotional and physical well-being at follow-up (Meister et al., 2015). Furthermore, patients’ coping styles, which can be defined as relatively permanent, individual-specific ways of facing difficulties in stressful situations, may play a role in this context (Du et al., 2016, Lazarus and Folkman, 1984). While coping originally was described as a dynamic process, most researchers over the past decades have operationalized coping as trait-like strategies. Coping styles or “defenses” are involuntary behaviors that are used to shield from sudden changes, in this case the cardiac event (Vaillant, 2000, Vaillant, 2011). There are several approaches to classify coping styles. They can be allocated according to level of maturity, where immature defenses include behaviors such as projection, while sublimation and humor are considered mature defenses (Vaillant, 2000, Vaillant, 2011). Another perspective distinguishes different categories of coping styles, such as emotion-focused (e.g., blame myself for procrastinating) and task-oriented coping (e.g., outline my priorities). Whether coping styles are adaptive or maladaptive depends on the situational context.

Most research on personality and coping styles is based on a variable-centered approach, which strives to group similar variables or “traits” together. This approach assumes that the population is homogeneous in how these traits operate on outcomes (Cooper and Larsen, 2013, Laursen and Hoff, 2006). A person-centered approach aims to identify groups of individuals who share particular attributes or show similar scoring “profiles” (Cooper and Larsen, 2013, Laursen and Hoff, 2006). A profile captures unique personality information that is not well covered by the use of multiple trait scores (Asendorpf, 2015). Identifying profiles based on relatively stable traits might be helpful for a better understanding of the heterogeneity in psychosocial characteristics that may affect the clinical course of CAD, but also for personalized medicine, which aims to individualize care according to the patients’ unique characteristics (Chapman et al., 2007, Denollet and Kupper, 2015, Fried, 2017).

The aim of the current study was to apply a person-centered approach to personality traits and coping styles based on latent class analysis in patients with CAD. We also examined the predictive value of this approach in explaining heterogeneity in patient-reported outcomes at six months follow-up.