Derivation and validation of a predictive model for chronic stress in patients with cardiovascular disease.

BACKGROUND: Chronic stress in patients with cardiovascular disease (CVD), including peripheral artery disease (PAD), is independently associated worse outcomes. A model that can reliably identify factors associated with risk of chronic stress in patients with CVD is needed.
METHODS: In a prospective myocardial infarction (MI) registry (TRIUMPH), we constructed a logistic regression model using 27 patient demographic, socioeconomic, and clinical factors, adjusting for site, to identify predictors of chronic stress over 1 year. Stress at baseline and at 1-, 6- and 12-month follow-up was measured using the 4-item Perceived Stress Scale (PSS-4) [range 0-16, scores ≥6 depicting high stress]. Chronic stress was defined as at least 2 follow-up PSS-4 scores ≥6. We identified and validated this final model in another prospective registry of patients with symptomatic PAD, the PORTRAIT study.
RESULTS: Our derivation cohort consisted of 4,340 patients with MI (mean age 59.1 ± 12.3 years, 33% females, 30% non-white), of whom 30% had chronic stress at follow-up. Of the 27 factors examined, female sex, current smoking, socioeconomic status, and economic burden due to medical care were positively associated with chronic stress, and ENRICHD Social Support Instrument (ESSI) score and age were inversely related to chronic stress. In the validation cohort of 797 PAD patients (mean age 68.6±9.7 years, 42% females, 28% non-white, 18% chronic stress) the c-statistic for the model was 0.77 and calibration was excellent.
CONCLUSIONS: We can reliably identify factors that are independently associated with risk of chronic stress in patients with CVD. As chronic stress is associated with worse outcomes in this population, our work identifies potential targets for interventions to as well as the patients that could benefit from these.

Background

High stress levels have been associated with development of cardiovascular disease (CVD) [1], and in patients with CVD, with adverse outcomes, including death [2,3], recurrent events [4,5] and poorer quality of life [2]. The latest American College of Cardiology and American Heart Association guidelines on prevention of CVD recommend addressing psychosocial stressors as a preventive measure to decrease cardiovascular risk [6]. Randomized controlled trials testing interventions to mitigate the impact of stress, including cognitive behavioral therapy [7], transcendental meditation [8], and group psychotherapy sessions [9] have demonstrated a decreased risk of death and recurrent events in patients with CVD. However, the results of these trials have not been widely adopted in clinical practice. Strategies to prevent worse outcomes in CVD could be enhanced by integration of interventions specifically targeting chronic stress into the broader context of cardiovascular care. A prerequisite to his goal is to understand the totality of patient and societal factors contributing to the risk of chronic stress in patients with CVD.

To our knowledge, no risk model has been validated to predict chronic stress in patients with CVD. Such a model is needed to understand the impact of factors that play a role in the development of chronic stress in patients with CVD and to identify patients who may benefit from future programs that address these underlying factors directly in conjunction with cardiovascular rehabilitation programs with integrated care pathways to help manage chronic stress. We aimed to develop such a model in a cohort of patients who survived acute myocardial infarction (AMI) and to validate the model in a cohort of patients with peripheral artery disease (PAD).

Methods

The investigators are willing to work with others, who are interested in validating or extending our analyses. For the PORTRAIT study data requests can be sent to Yale University Institute Review Board, at hrpp@yale.edu. For the TRIUMPH study data requests could be sent to the steering committee for the TRIUMPH registry at dbuchanan@saint-lukes.org.

Study population

Data from a prospective registry of patients presenting with AMI, the Translational Research Investigating Underlying disparities in acute Myocardial infarction Patient’s Health Status (TRIUMPH) study was used for model derivation [10]. For external validation of our model, we used data from a prospective registry of patients presenting with worsening Peripheral Artery Disease (PAD), the Patient-centered Outcomes Related to Treatment Practices in Peripheral Arterial Disease: Investigating Trajectories (PORTRAIT) registry [11]. Although both TRIUMPH and PORTAIT included patients who had a different presentation of CVD, both included a patient population who had worsening of their CVD and examined trajectories of stress using the same instrument to quantify stress, social support, as well as similar patient demographic, psychosocial, clinical and socioeconomic measures.

Registry designs

The design of both the TRIUMPH and PORTRAIT study have been published elsewhere [10,11]. The TRIUMPH study enrolled patients (n = 4,340) presenting with MI who were alive at hospital discharge. Enrollment was done form April 11, 2005 to December 31, 2008 across 24-US hospitals. Patients who were enrolled had biomarker evidence of myocardial necrosis and additional clinical evidence supporting the diagnosis of AMI, including prolonged ischemic signs/symptoms or electrocardiographic criteria of ST segment changes. Baseline data were obtained through chart abstraction and structured interviews by trained research coordinators. Data on health status and psychosocial stress were obtained at baseline, 1-, 6- and 12-month follow-up using a standardized interview conducted by trained study personnel.

The PORTRAIT registry enrolled patients who presented with worsening symptoms of PAD to sub-specialty clinics in the US (n = 797), Australia (n = 95) and Netherlands (n = 383) from June 2011 to December 2015. For this study, only the patients from the US were included. Enrolled patients had an ankle brachial index (ABI) ≤ 0.90 or a significant drop in post-exercise ankle pressure (≥ 20mm of Hg). Patient demographics, health status, psychosocial characteristics, socioeconomic variables, and cardiovascular lifestyle factors were obtained through interviews at the initial visit. Patient symptoms, medical history, comorbidities, and PAD diagnostic information were abstracted from medical records. Serial information about health status and patient psychosocial profile was collected at baseline, 3-, 6- and 12-month follow-up through centralized follow-up. For both TRIUMPH and PORTRAIT studies, all study participants provided written or telephonic informed consent and the study protocol was approved by Institution Review Boards of Saint Luke’s Hospital and all participating sites.

Assessment of stress and definition of chronic stress

In both TRIUMPH and PREMIER, level of perceived stress was assessed at enrollment and follow-up with the 4-item perceived stress scale (PSS-4). The PSS-4 is a reliable and valid measure (Cronbach’s Alpha 0.67–0.79) of an individual’s self-evaluation of control and confidence in handling the stressful situations they have experienced over the past month [12]. In our study, the Cronbach’s Alpha was 0.79 for the TRIUMPH cohort and 0.68 for the PORTRAIT cohort. Scores on the PSS-4 range from 0–16, with higher scores indicating higher stress and lower ability to cope with that stress [12]. The PSS-4 is a non-diagnostic instrument and there are no established thresholds, although in patients with cardiovascular disease, a score of ≥6 has been associated with adverse outcomes in patients after MI [2]. Hence, in keeping with prior research we used a score of ≥6 as the threshold to describe high levels of perceived stress. PSS-4 was collected at baseline and at each follow-up assessment. We wanted to quantify a patient’s exposure to chronic stress, during the 12-months of follow-up. Therefore, to provide more stable categorization of stress levels at follow-up we defined chronic stress as 2 or more follow-up PSS-4 assessments of ≥6, after the initial baseline assessment. As the initial event (AMI or worsening PAD symptoms) could contribute to the patient’s stress in the first few days, baseline PSS-4 assessments were not included in the definition.

Assessment of socioeconomic status

General socioeconomic status (SES) was assessed using the question, “how much money do you have left over at the end of the month?” with possible responses being “enough”, “just enough” and “not enough”. Economic burden due to medical care was assessed using the question, “What is the economic burden of your medical costs” with possible responses being “severe burden”, “moderate burden”, “somewhat of a burden”, “a little burden” and “no burden at all”.

Social support and disease-specific health status

Social support was quantified using the ENRICHD Social Support Instrument (ESSI), which has been derived from the Medical Outcomes Survey and prior work examining the influences of social support [13]. The ESSI is a 7-item measure and assesses four attributes of social support: emotional, instrumental, informational and appraisal [13,14]. The ESSI was found to be a valid and reliable measure of social support in patients to screen for patients enrolled in a depression intervention trial [15].

Statistical analysis

Patient demographic, socioeconomic, clinical factors and PSS-4 scores at baseline were described separately in our derivation and validation cohorts. To identify predictors of chronic stress over 1 year, we constructed a multivariable hierarchical logistic regression on all patient factors (listed in Table 1), adjusting for study site as a random effect. No appreciable multicollinearity was found among the predictors (all variance inflation factors < 2.0; design matrix condition index 17.6). Nonlinear effects for continuous variables were examined using restricted cubic splines; however, no significant nonlinearity was detected (p = 0.73), so effects were refit linearly for parsimony and ease of interpretation. There was moderate site-level variability in the outcome of chronic stress over 1-year (Median Odds Ratio 1.21), and to account for this variability we added site as a random effect in the model. Model performance was assessed using the c-statistic to determine discrimination and by plotting deciles of predicted risk against the observed event rate and comparing the regression line with the line of unity (intercept = 0 and slope = 1). Finally, to understand the prevalence of independent predictors of chronic stress in patients stratified by age, we compared the prevalence of predictors in patients <55 and ≥ 55 years of age.

Table 1

Baseline patient characteristics and prevalence of chronic stress in patients from the TRIUMPH study.

	Total n (%)	Prevalence of Chronic Stress	p-value
Demographics
Age (years)    19.0 to <55    55 to <65    65 to <75    75 to 98.0	1633 (37.6%)1374 (31.7%)782 (18.0%)551 (12.7%)	40.4%31.8%17.8%19.5%	<0.001
MaleFemale	2898 (66.8%)1442 (33.2%)	28.1%36.7%	<0.001
Non-whiteWhite	1305 (30.2%)3022 (69.8%)	38.0%27.9%	<0.001
Socioeconomic Factors
MarriedNot Married	2318 (53.5%)2014 (46.5%)	24.6%38.2%	<0.001
Finances at the End of the Month    Some money left over    Just enough to make ends meet    Not enough to make ends meet	1777 (41.7%)1592 (37.4%)889 (20.9%)	15.7%34.7%54.6%	<0.001
Not WorkingWorking	2200 (51.2%)2100 (48.8%)	34.8%26.9%	<0.001
Education    Less than high school    High school    College degree	895 (20.7%)2542 (58.9%)878 (20.3%)	38.9%31.3%21.8%	<0.001
Has avoided care due to costHas not avoided care due to cost	1088 (25.6%)3165 (74.4%)	51.2%24.0%	<0.001
Medical Costs Economic Burden    Severe burden    Moderate burden    Somewhat of a burden    A little burden    No burden at all	447 (10.5%)409 (9.6%)507 (11.9%)440 (10.3%)2462 (57.7%)	55.1%46.4%38.9%30.3%22.4%	<0.001
Lives aloneDoes not live alone	1061 (24.6%)3247 (75.4%)	34.5%29.8%	0.023
ESSI score    5 to <20    20 to <25    25 to 25	886 (21.1%)1328 (31.6%)1985 (47.3%)	52.0%28.9%22.8%	<0.001
Comorbid Medical Conditions
Body Mass Index (kg/m2)    13.5 to <25    25 to <30    30 to <35    35 to 70.7	959 (23.3%)1467 (35.7%)974 (23.7%)709 (17.3%)	29.6%27.2%32.1%39.0%	0.001
Smoking Status    Current    Former    Never	1689 (39.2%)1403 (32.6%)1215 (28.2%)	41.3%22.8%25.9%	<0.001
Hypertension    Yes    No	2893 (66.7%)1447 (33.3%)	32.4%28.1%	0.008
Diabetes    Yes    No	1336 (30.8%)3004 (69.2%)	34.9%29.2%	0.006
Dyslipidemia    Yes    No	2128 (49.0%)2212 (51.0%)	30.2%31.7%	0.37
Prior Percutaneous Coronary Intervention    Yes    No	851 (19.6%)3489 (80.4%)	33.2%30.4%	0.18
Prior Coronary Artery Bypass Graft Surgery    Yes    No	495 (11.4%)3845 (88.6%)	31.2%30.9%	0.93
Prior Myocardial Infarction    Yes    No	912 (21.0%)3428 (79.0%)	33.0%30.4%	0.20
Prior stroke/transient ischemic attack    Yes    No	304 (7.0%)4036 (93.0%)	33.5%30.8%	0.40
Congestive Heart Failure    Yes    No	372 (8.6%)3968 (91.4%)	39.9%30.1%	0.002
Atrial Fibrillation    Yes    No	212 (4.9%)4128 (95.1%)	32.4%30.9%	0.70
Chronic Kidney Disease    Yes    No	322 (7.4%)4018 (92.6%)	31.4%30.9%	0.88
PHQ-8 Depression Score    0.0 to <5    5 to <10    10 to <15    15 to 24.0	2322 (56.8%)1009 (24.7%)464 (11.3%)294 (7.2%)	18.8%35.0%55.6%70.8%	<0.001
Baseline stress (PSS-4 ≥ 6)    Yes    No	1622 (38.6%)2582 (61.4%)	54.2%16.3%	<0.001
Hospital Presentation
Non-ST elevation MI    Yes    No	2473 (57.0%)1867 (43.0%)	33.0%28.2%	0.003
In-Hospital Revascularization    None    PCI    CABG	1153 (26.6%)2782 (64.1%)405 (9.3%)	37.8%28.5%28.0%	<0.001

ESSI = ENRICHD social support index, PHQ-8 = 8-point Patient Health Questionnaire depression scale, SAQ = Seattle Angina Questionnaire.

Missing data

Of the 4,340 patients in TRIUMPH, 1,682 had complete PSS-4 scores at three follow-up assessments, 1,105 at only two assessments, 786 at only one assessment, and 767 had no follow-up scores. Scores were missing due to skipped items (2.8%), refusals (4.1%), illness (2.4%), lost to follow-up (24.8%) or death (4.0%). We used multiple imputation by chained equations (MICE) with predictive mean matching to impute missing PSS-4 scores (as well as missing values of candidate predictor variables) [16]. The imputation model included all available PSS-4 questionnaire items at all time points, site as well as the 27 predictors of interest that we identified a priori based on previous literature and clinical judgement. A total of 20 randomly imputed data sets were generated, and the outcome of chronic stress was defined on each data set using observed and imputed scores. All analyses were performed by analyzing each of the 20 data sets separately and then pooling the results, to account for bias and uncertainty due to missingness.

Results

Patient populations

In the derivative cohort (TRIUMPH study), the mean age of the study population was 59.1±12.3 years, 33.2% were females and 30.2% were non-white. Overall, 30% of the patients had chronic stress at follow-up. Table 1 describes baseline patient characteristics for the candidate variables in the derivation cohort and compares prevalence of chronic stress among patients stratified by their baseline characteristics. Patients in age groups of 19–54 years had the highest prevalence of chronic stress. Moreover, prevalence of chronic stress was higher in non-whites, in patients who reported not having enough finances at month’s end, and in patients who perceived healthcare costs to be a severe economic burden.

In the validation cohort (PORTRAIT study), the mean age was 68.6 ± 9.7 years, 41.9% of the patients were females and 27.6% were non-white. Overall 18% of the patients had chronic stress at follow-up. Table 2 describes the baseline patient characteristics and prevalence of chronic stress among patients stratified by each baseline factor. Patients aged 42–54 years, females, non-whites, and patients who reported end of month financial distress and severe burden of healthcare costs had a higher prevalence of chronic stress.

Table 2

Baseline patient characteristics and prevalence of chronic stress in patients from the PORTRAIT study.

	Total n (%)	Prevalence of Chronic Stress	p-value
Demographics
Age (years)    42.0 to <55    55 to <65    65 to <75    75 to 94.8	61 (7.7%)202 (25.3%)311 (39.0%)223 (28.0%)	46.5%27.4%11.5%12.1%	<0.001
MaleFemale	463 (58.1%)334 (41.9%)	14.9%23.2%	0.007
Non-whiteWhite	220 (27.6%)577 (72.4%)	25.5%15.6%	0.006
Socioeconomic Factors
MarriedNot Married	435 (55.0%)356 (45.0%)	14.9%22.5%	0.016
Finances at the End of the Month    Some money left over    Just enough to make ends meet    Not enough to make ends meet    Missing	406 (51.4%)291 (36.8%)93 (11.8%)7	9.0%24.9%38.5%	<0.001
Not WorkingWorking	613 (77.5%)178 (22.5%)	19.7%13.7%	0.13
Education    High school    College degree    Missing	499 (73.7%)178 (26.3%)120	32.6%17.6%10.8%	0.001
Has avoided care due to costHas not avoided care due to cost	130 (16.4%)661 (83.6%)	35.2%15.0%	<0.001
Medical Cost Economic Burden    Severe burden    Moderate burden    Somewhat of a burden    A little burden    No burden at all	31 (3.9%)71 (9.0%)86 (10.9%)95 (12.0%)509 (64.3%)	46.2%29.8%30.8%16.4%13.2%	<0.001
Lives aloneDoes not live alone	210 (26.4%)586 (73.6%)	19.9%17.8%	0.60
ESSI score    5 to <20    20 to <25    25 to 25	139 (17.6%)221 (27.9%)432 (54.5%)	37.2%14.9%14.0%	<0.001
Comorbid Medical Conditions
BMI    15.2 to <25    25 to <30    30 to <35    35 to 60.5	186 (24.2%)270 (35.1%)192 (24.9%)122 (15.8%)	18.7%16.1%19.5%21.1%	0.46
Smoke status    Never    Former    Current	104 (13.1%)451 (56.7%)241 (30.3%)	15.1%14.4%27.1%	0.003
Hypertension    Yes    No	706 (88.6%)91 (11.4%)	18.3%18.5%	0.99
Diabetes    Yes    No	305 (38.3%)492 (61.7%)	20.6%17.0%	0.34
Dyslipidemia    Yes    No	706 (88.6%)91 (11.4%)	17.8%22.6%	0.38
Prior Revascularization Procedure    Yes    No	358 (44.9%)439 (55.1%)	18.3%18.4%	0.99
Prior Myocardial Infarction    Yes    No	176 (22.1%)621 (77.9%)	21.4%17.5%	0.31
Prior stroke/transient ischemic attack    Yes    No	93 (11.7%)704 (88.3%)	22.2%17.8%	0.42
Congestive heart failure    Yes    No	115 (14.4%)682 (85.6%)	22.7%17.6%	0.32
Atrial fibrillation    Yes    No	108 (13.6%)689 (86.4%)	18.7%18.3%	0.97
Chronic kidney disease    Yes    No	121 (15.2%)676 (84.8%)	21.6%17.8%	0.47
PHQ-8 Depression Score    0.0 to <5    5 to <10    10 to <15    15 to 24.0	477 (61.5%)159 (20.5%)81 (10.5%)58 (7.5%)	7.3%22.2%41.3%64.8%	<0.001
Baseline stress (PSS-4 ≥ 6)    Yes    No	282 (35.8%)506 (64.2%)	34.4%9.4%	<0.001

ESSI = ENRICHD social support index, PHQ-8 = 8-point Patient Health Questionnaire depression scale, PAQ = Peripheral Artery Disease Questionnaire.

Predictive model

Fig 1 describes the baseline patient factors evaluated in the model. Of all the baseline patient factors examined (Fig 1), 6 were found to be independently associated with outcome of chronic stress over 12-month follow-up. S1 Fig details the calibration plot for the 6-item model in the TRIUMPH study. These were age, sex, economic burden related to medical care, general SES, current smoker, and ESSI score. The bootstrapped-validated c-statistic for the final model (including all the 6-items) was 0.75 (S1 Fig). The c-statistic for the final model, applied to the validation cohort (PORTRAIT study) was 0.77 and calibration was excellent (Fig 2). S1 Table gives the intercept and coefficient information for all the covariates in the final regression equation.

Fig 1

Independent predictors of chronic stress in the TRIUMPH study.

Fig 2

External validation of the predictive model in the PORTRAIT study.

Prevalence of predictors of chronic stress in patients stratified by age

Table 3 shows the prevalence of predictors of chronic stress in our derivative cohort, stratified by age. The socioeconomic predictors of chronic stress were more prevalent in patients younger than 55 years old compared with older patients.

Table 3

Socioeconomic predictors of chronic stress in patients stratified by age, in TRIUMPH study.

Socioeconomic Predictors of Chronic Stress
	<55 yearsn = 1,633	>55 yearsn = 2,707	p-value
Finances at the End of the Month    Some money left over    Just enough to make ends meet    Not enough to make ends meet	546 (33.9%)618 (38.3%)448 (27.8%)	1231 (46.5%)974 (36.8%)441 (16.7%)	<0.001
Not WorkingWorking	540 (33.4%)1078 (66.6%)	1660 (61.9%)1022 (38.1%)	<0.001
Education    Less than high school    High school    College degree	313 (19.2%)1047 (64.4%)266 (16.4%)	582 (21.6%)1495 (55.6%)612 (22.8%)	<0.001
Has avoided care due to costHas not avoided care due to cost	555 (34.6%)1051 (65.4%)	555 (34.6%)1051 (65.4%)	<0.001
Medical Costs Economic Burden    Severe burden    Moderate burden    Somewhat of a burden    A little burden    No burden at all	219 (13.6%)175 (10.9%)219 (13.6%)136 (8.5%)858 (53.4%)	228 (8.6%)234 (8.8%)288 (10.8%)304 (11.4%)1604 (60.3%)	<0.001
Lives aloneDoes not live alone	338 (20.8%)1288 (79.2%)	723 (27.0%)1959 (73.0%)	<0.001
ESSI score    5 to <20    20 to <25    >25	378 (23.8%)500 (31.5%)709 (44.7%)	508 (19.4%)828 (31.7%)1276 (48.9%)	0.001

Discussion

Observational studies describing the adverse cardiovascular outcomes due to chronic stress exposure have been described since the 1970s and have been replicated in diverse clinical settings and in patients across a spectrum of CVD risk [2,17-20]. Mechanisms that could explain the development and progression of CVD due to chronic stress have also been described and include direct pathophysiological effects [21] and indirect pathways through adverse health behaviors [22]. Chronic stress as a construct is germane to individual patient and societal influences and understanding the impact of these factors in its totality is important. We found that 30% of the patients who experience an AMI and 18% of the patients diagnosed with symptomatic PAD continue to suffer from chronic stress over 1-year of follow-up. This was after the initial event (AMI, worsening symptoms of PAD), which could contribute to stress in the first few days. We developed a risk model in these contemporary cohort of patients with CVD to understand the role of patient and environmental factors towards development of chronic stress. These insights could be invaluable for designing and testing novel stress-reduction strategies in at-risk patient populations and to prioritize preventive policies and programs that could help address systemic sources of distress.

We identified 6 predictors of chronic stress in our CVD cohorts, which were age, sex, economic burden related to medical care, general SES, current smoker, and ESSI score. These predictors have been associated with stress in more heterogeneous populations. For example, financial strain has been associated with chronic stress and adverse outcomes [23,24], and age has also been shown to influence chronic stress, though the literature here is inconsistent. Some studies have demonstrated that older individuals are less affected by environmental stressors [25,26], while others have demonstrated that they are more vulnerable [27] or have found no association of stress with age [28]. Cognitive theories of aging have postulated that older adults use attentional strategies and reappraisals more frequently to mitigate the impact of environmental influences to avoid chronic stress [29,30]. Moreover, the prevalence of other socioeconomic predictors of chronic stress was higher in younger adults, which could also explain our findings. Smoking has been associated with high stress in several previous studies [31,32]. While the self-medication hypothesis for stress and smoking may explain this link, it is also known that smoking is marker of a higher likelihood of chronic stress [33]. This finding was reflected in our results. Furthermore, lack of a social network structure has been linked to anxiety [34] and social support has been shown to buffer the impact of stress on the wellbeing of individuals facing financial hardship [35,36].

From a public health standpoint, it is important to highlight that while public health policies target population health directly through measures such as screening guidelines, immunization etc., social policies to improve SES have also been shown to improve health of the population. For example, in the United Kingdom, implementation of National Minimum Wage legislation in 1999 was associated with improved mental health of low-wage workers [37]. In the US, improvement in SES via programs such as Social Security have had a beneficial impact on the health of the elderly [38]. There is evidence that unconditional universal basic income has positive effect on population health outcomes [39]. In Scotland for example, policies such as citizens basic income were associated with improved population health outcomes [40]. SES, economic burden of health care, and lack of social support were strong predictors of chronic stress in both of our CVD cohorts. Economic plans such as universal basic income, universal coverage of health care, and expansion of social security which would improve SES and the economic burden of medical care for patients with CVD, should be tested to assess its impact as a stress-reduction strategy in patients with CVD.

Regardless of its trigger or root cause, experiencing chronic stress is linked with an increased cardiovascular risk. It is also known, however, that stress is a modifiable risk factor for which evidence-based management strategies exist [41]. Equipping patients with coping skills to reduce stress in their lives has been shown to be effective in improving quality of life in patients with coronary artery disease [21]. Furthermore, chronic stress management through cognitive behavioral therapy programs [7], as well as through transcendental meditation [8], in addition to standard care, has been shown to reduce the risk of recurrent cardiovascular events in patients with coronary artery disease and to prolong women’s life following an acute myocardial infarction. Given the strength of the association found in our study and the fact that this risk factor has been largely ignored in the CVD population, there is an important need for future studies to test the efficacy of stress management strategies on cardiovascular outcomes. The current work can help identify patients who would be most likely to benefit from such interventions.

Our study also had some limitations. Both derivation and validation data sets were obtained from prospective registries that included carefully selected institutions. Whether the enrolling institutes for both TRIUMPH and PORTRAIT studies are representative of other sites not included in these studies is not known. Second, societal, cultural and sociopolitical influences are unique to the US cohorts under study, and whether the major predictors of chronic stress are similar in other countries remains an important area of further work. Third, we quantified stress using the PSS-4 which is a generic and brief instrument to assess perceived stress levels in communities and this measurement may not necessarily extend to other measures of stress or other domains of mental health functioning, nor should it be used for diagnosing purposes, as stress reactions are universal responses. Fourth, the derivative and validation cohorts differed in terms of demographics, socioeconomic conditions and vascular disease (coronary artery disease vs PAD). However, both disease processes are a manifestation of the same pathophysiological mechanism of atherosclerosis. Moreover, patients with PAD have a similar risk of adverse cardiovascular events (myocardial infarction, stroke), compared to patients diagnosed with coronary artery disease [42]. Our model had good predictive ability in both cohorts, underscoring the value of our model in screening for higher levels of stress in patients across the spectrum of cardiovascular disease. Fifth, it is known that factors at the workplace are associated with risk of experiencing chronic stress and have been associated with development and progression of CVD [1]. Stress at the work-place along with other factors such as economic hardship, lack of social support etc., are important sources of stress for patients with CVD [43,44]. Our aim was to identify patients who are at risk of CVD, and we did not further explore individual sources of stress. Indeed, identifying unique sources of stress in patients with CVD, with the aim of formulating actionable coping strategies, in addition to identifying policy measures that could address some of the more systematic root causes of stress remain areas for future work Finally, approximately 25% of patients had missing follow-up PSS-4 scores due to loss to follow-up, which is similar to missing follow-up rates seen in other prospective AMI registries [45]. We used multiple imputation to account for missing data, but there remains potential for bias. However this still remains an important limitation of our work”.

Conclusion

A substantial number of patients with CVD suffer from chronic stress. We describe and externally validated a well performing prediction model that prognosticates the risk of chronic stress in patients with CVD. As exposure to chronic stress has been linked to adverse clinical outcomes in this population, our model provides valuable insights into the identification of patient-level and societal predictors that could inform design and testing of preventive programs in conjunction with targeted stress-reduction strategies and the patients that may benefit from these in the future.

Calibration plot for the 6-item model in the TRIUMPH study.

(JPG)

Click here for additional data file.

Model coefficients for all factors in the final model to predict chronic stress in patients with cardiovascular disease.

(DOCX)

Click here for additional data file.

17 Mar 2022

Derivation and Validation of a Predictive Model for Chronic Stress in Patients with Cardiovascular Disease

PLOS ONE

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

M. Harvey Brenner, PhD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for including your ethics statement:  "For both TRIUMPH and PORTRAIT studies, all study participants provided written or telephonic informed consent and the study protocol was approved by Institution Review Boards of Saint Luke’s Hospital and all participating sites."

Please amend your current ethics statement to include the full name of the ethics committee/institutional review board(s) that approved your specific study.

Once you have amended this/these statement(s) in the Methods section of the manuscript, please add the same text to the “Ethics Statement” field of the submission form (via “Edit Submission”).

For additional information about PLOS ONE ethical requirements for human subjects research, please refer to http://journals.plos.org/plosone/s/submission-guidelines#loc-human-subjects-research.

3. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match.

When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section.

4. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

5. Please note that in order to use the direct billing option the corresponding author must be affiliated with the chosen institute. Please either amend your manuscript to change the affiliation or corresponding author, or email us at plosone@plos.org with a request to remove this option.

6. We noticed you have some minor occurrence of overlapping text with the following previous publication(s), which needs to be addressed:

- https://mospace.umsystem.edu/xmlui/handle/10355/71054

- https://linkinghub.elsevier.com/retrieve/pii/S0022399920308758

In your revision ensure you cite all your sources (including your own works), and quote or rephrase any duplicated text outside the methods section. Further consideration is dependent on these concerns being addressed.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: It is of course elegant to establish the prediction model in one cohort and then apply it on another one. However, I feel that the authors have not discussed the significance of the obvious differences between the two cohorts. The PAD cohort is older and among participants only one fourth are working whereas in the TRIUMPH CAD cohort the participants are younger and half them are working. Coronary heart disease is associated with threat to life whereas PAD is associated with pain suffering and severe limitations in movement ability. However for those who are working (in both groups) working conditions are an important potential source of chronic stress. A substantial part of the literature on psychosocial prediction of CAD is based upon stress at work studies. This pertains for instance to the review by Kivimäki and Steptoe that the authors refer to. In fact most of the studies that they refer to deal with stress at work. In addition the international IPD study fairly recently reported from a prospective study that job strain significantly predicts development of PAD (Heikkilä et al 2020). It is odd to see the complete lack of reference to this whole literature.

In view of the great importance of working conditions you might in the future use one questionnaire for working and a slightly different one for a non-working group.

I can see a point in using the short four-item stress questionnaire for screening purposes, with subsequent more detailed discussion regarding working conditions in individual cases. The authors mention that the questionnaire captures factors associated with both environmental and individual (coping) factors. For less informed readers it may be helpful to explain that follow-up interviews may help in determining a strategy for dealing with the patient´s chronic stress and that things to do might have to do either with coping strategies or family/work organization questions.

The authors hint at the possibility that chronic stress during the follow-up period might have been a consequence of medical procedures and changes in disease course. I think that point needs to be developed because it would be of interest to clinicians reading the paper. Conversely, another topic the authors do not discuss is the potential role of adverse psychosocial conditions in aggravating the clinical course

Despite your professional efforts to explore effects of drop-out on different levels I still think you should discuss more openly the fact that a lot of subjects have fallen behind in several drop-out steps. The ones remaining are likely to be under less chronic stress than others.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Tores Theorell

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

3 Jun 2022

Please note: Below, we have copied the reviewers’ comments (in bold) and provided our responses in plain text. Sentences added to the manuscript in response to an editor or reviewer comment are indented and placed in italic font. ________________________________________

Reviewer Comments:

It is of course elegant to establish the prediction model in one cohort and then apply it on another one. However, I feel that the authors have not discussed the significance of the obvious differences between the two cohorts. The PAD cohort is older and among participants only one fourth are working whereas in the TRIUMPH CAD cohort the participants are younger and half them are working. Coronary heart disease is associated with threat to life whereas PAD is associated with pain suffering and severe limitations in movement ability.

RESPONSE# 1: These are important points. The reviewer is correct in pointing out the differences between the derivative cohort (TRIUMPH study) and validation cohort (PORTRAIT study), with the obvious difference of vascular bed involved, coronary in TRIUMPH cohort and peripheral in PORTRAIT cohort.

From a pathophysiological standpoint, the two conditions coronary artery disease (CAD) and peripheral artery disease (PAD) have the same underlying etiology, i.e., atherosclerosis as a generalized process. Even though the manifestation of symptoms is different, prior work has demonstrated that the clinical event burden in terms of future risk of myocardial infarction and stroke, is similar, if not greater amongst those with PAD vs. CAD (JAMA. 2007 Mar 21;297(11):1197-206). So even though the cohorts represent different arterial beds that got affected, they all represent manifestations of the same underlying disease. It is reassuring to know that the performance of the models across these populations is comparable, which would accommodate screening for high levels of chronic stress throughout atherosclerotic disease trajectories that patients may face. This point is underscored by the c-statistic for our 6-item model which was comparable in both cohorts, 0.75 in TRIUMPH cohort and 0.77 in the PORTRAIT cohort.

The following sentences have been added to the discussion section.

Page 12 Line 22-23, Page 13 Line 1-5

“Fourth, the derivative and validation cohorts differed in terms of demographics, socioeconomic conditions and vascular disease (coronary artery disease vs PAD). However, both disease processes are a manifestation of the same pathophysiological mechanism of atherosclerosis. Moreover, patients with PAD have a similar risk of adverse cardiovascular events (myocardial infarction, stroke), compared with patients affected by coronary artery disease.42 Our model had good predictive ability in both cohorts, underscoring the value of our model in screening for higher levels of stress in patients across the spectrum of cardiovascular disease. “

However for those who are working (in both groups) working conditions are an important potential source of chronic stress. A substantial part of the literature on psychosocial prediction of CAD is based upon stress at work studies. This pertains for instance to the review by Kivimäki and Steptoe that the authors refer to. In fact most of the studies that they refer to deal with stress at work. In addition the international IPD study fairly recently reported from a prospective study that job strain significantly predicts development of PAD (Heikkilä et al 2020). It is odd to see the complete lack of reference to this whole literature. In view of the great importance of working conditions you might in the future use one questionnaire for working and a slightly different one for a non-working group.

RESPONSE# 2: We agree with the reviewer. Certainly, some of the literature exploring the impact of stress on development and progression of cardiovascular disease pertains to stress at the workplace. We have acknowledged this in our revision and referred to the work by Heikkilä et al.

It is indeed important to understand sources of stress for patients with cardiovascular disease. Stress at workplace is indeed an important consideration. However other sources of stress such as economic hardship, lack of social support also contribute to the overall burden of stress in these patients (J Am Coll Cardiol. 2018 Jun 5;71(22):2585-2597, Heart. 2020 Sep;106(18):1394-1399). Future work should focus on identifying these causes of stress to help identify interventions to help patients cope, as a strategy to improve outcomes, in addition to identifying policy measures that could address some of the more systematic root causes of stress. While we did not explore source of stress with aim of formulating actionable coping mechanisms, this remains an area for future work.

The following sentences have been added to the discussion section.

Page 13 Line 5-13

“Fifth, it is known that factors at the workplace are associated with risk of experiencing chronic stress and have been associated with development and progression of CVD.1 Stress at the work-place along with other factors such as economic hardship, lack of social support etc., are important sources of stress for patients with CVD.43,44 Our aim was to identify patients who are at risk of CVD, and we did not further explore individual sources of stress. Indeed, identifying unique sources of stress in patients with CVD, with the aim of formulating actionable coping strategies, in addition to identifying policy measures that could address some of the more systematic root causes of stress remain areas for future work”.

I can see a point in using the short four-item stress questionnaire for screening purposes, with subsequent more detailed discussion regarding working conditions in individual cases. The authors mention that the questionnaire captures factors associated with both environmental and individual (coping) factors. For less informed readers it may be helpful to explain that follow-up interviews may help in determining a strategy for dealing with the patient´s chronic stress and that things to do might have to do either with coping strategies or family/work organization questions.

RESPONSE# 3: This is a very important point. As next steps, we will need to further identify the sources of stress patients affected by cardiovascular disease are dealing with (e. g. work stress, loneliness, economic hardship, as well as stress from navigating the disease and its management itself etc.) and develop tailored interventions to manage or mitigate sources of stress. This remains an area of future work. Please see RESPONSE# 2.

The authors hint at the possibility that chronic stress during the follow-up period might have been a consequence of medical procedures and changes in disease course. I think that point needs to be developed because it would be of interest to clinicians reading the paper. Conversely, another topic the authors do not discuss is the potential role of adverse psychosocial conditions in aggravating the clinical course.

RESPONSE# 4: The reviewer is correct. We were cognizant of this issue and agree that stress during follow-up could be due to medical procedures and changes in disease course. To account for this bias, we excluded baseline stress assessment from the definition of chronic stress. We had mentioned that in the methods section and have clarified this further in the revised version of the manuscript. For TRIUMPH study (derivation cohort), follow-up assessments were at 1,6- and 12-month follow-up intervals. For the PORTRAIT study (validation cohort) follow-up assessments were at 3,6 and 12-month follow-up intervals. We defined chronic stress as two or more follow-up assessments with PSS-4 score ≥ 6.

The following sentences were added/modified to the manuscript.

Page 6 Line 2-7

“We wanted to quantify a patient’s exposure to chronic stress, during the 12-months of follow-up. Therefore, to provide more stable categorization of stress levels at follow-up we defined chronic stress as 2 or more follow-up PSS-4 assessments of �  6, after the initial baseline assessment. As the initial event (AMI or worsening PAD symptoms) could contribute to the patient’s stress in the first few days, baseline PSS-4 assessments were not included in the definition.”

Page 10 Line 10-11

“This was after the initial event (AMI, worsening symptoms of PAD), which could contribute to stress in the first few days.”

Despite your professional efforts to explore effects of drop-out on different levels I still think you should discuss more openly the fact that a lot of subjects have fallen behind in several drop-out steps. The ones remaining are likely to be under less chronic stress than others.

RESPONSE# 5: Thank you for the suggestion. We agree that missing data is a big limitation. Even though we used multiple imputations by chained equations (MICE) [Journal of statistical software 2010: 1-68], with predictive mean matching to account for any bias, this remains an important limitation of our work. We have highlighted this further in the limitations section of our revised manuscript.

The following sentences have been added to the manuscript.

Page 13 Line 15-16

“We used multiple imputation to account for missing data, but there remains potential for bias. However, this still remains an important limitation of our work”

Submitted filename: Response Letter v6.docx

Click here for additional data file.

23 Sep 2022

Derivation and Validation of a Predictive Model for Chronic Stress in Patients with Cardiovascular Disease

PONE-D-21-13011R1

Dear Dr.  Smolderen  

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Xianwu Cheng, M.D., Ph.D., FAHA

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Authors have adressed all original concerns.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors have been honest about weaknesses and discuss them adequately. They do contribute to the literature

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Tores Theorell

**********

7 Oct 2022

PONE-D-21-13011R1

Derivation and Validation of a Predictive Model for Chronic Stress in Patients with Cardiovascular Disease

Dear Dr. Smolderen:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Associate Prof. Xianwu Cheng

Academic Editor

PLOS ONE