European Journal of Preventive Cardiology

Abstract Background Increased pulse pressure is associated with structural target organ damage, especially in elderly patients, increasing cardiovascular risk. Design In this analysis, we investigated whether high pulse pressure retains a prognostic effect also when common markers of target organ damage are taken into account. Methods We analysed an unselected cohort of treated hypertensive patients from the Campania Salute Network registry (n = 7336). Participants with available cardiac and carotid ultrasound were required to be free of prevalent cardiovascular disease, with ejection fraction ≥50%, and no more than stage III Chronic Kidney Disease. The median follow-up was 41 months and end-point was occurrence of major cardiovascular events (i.e. fatal and non-fatal stroke or myocardial infarction and sudden death). Based on current guidelines, pulse pressure ≥60 mm Hg was classified as high pulse pressure (n = 2356), at the time of the initial visit, whereas pulse pressure <60 mm Hg was considered normal (n = 4980). Results High pulse pressure patients were older, more likely to be women and diabetic, while receiving more antihypertensive medications than normal pulse pressure (all p < 0.0001). High pulse pressure exhibited greater prevalence of left ventricular hypertrophy, and carotid plaque than normal pulse pressure (all p < 0.0001). In Cox regression, high pulse pressure patients had 57% increased hazard of major cardiovascular events, compared to normal pulse pressure (hazard ratio = 1.57; 95% confidence interval: 1.12–2.22, p = 0.01), an effect that was independent of significant prognostic impact of older age, male sex, diabetes, left ventricular hypertrophy, carotid plaque and less prescription of anti-renin–angiotensin system therapy. Conclusions High pulse pressure is a functional marker of target organ damage, predicting cardiovascular events in hypertensive patients, even independently of well-known structural markers of target organ damage. Hypertension, target organ damage, pulse pressure, cardiovascular events, arterial stiffness Introduction With advancing age, conduit vessels decrease their compliance and elasticity due to elastin degeneration, and collagen deposition.1 As a consequence, left ventricular (LV) pulsatile load increases, with substantial burden especially in patients with left ventricular dysfunction.2 Variation of pulsatile load may be clinically appreciated as pulse pressure (PP). Thus, the structural changes beyond the increase in PP indicate that high PP is a surrogate functional sign of target organ damage (TOD)3,4 in elderly patients, associated with increased cardiovascular (CV) events.4 Specifically, estimated central pulse pressure has been reported to be independently associated with adverse CV outcome in the cohort of the Strong Heart Study.5 More recent evidence suggests that increased brachial PP is associated with adverse CV outcome, also independently of mean arterial pressure, in high-risk patients with atherosclerosis6 and in patients with type 2 diabetes.7 However, whether high PP influences outcome also independently of structural markers of TOD, such as LV hypertrophy (LVH) and carotid atherosclerosis, is unknown. Accordingly, we evaluated whether increased PP impacts on incident CV events even independently of LVH and carotid plaque, in a large registry of hypertensive patients. Methods Patient population The Campania Salute Network (CSN) is an open electronic registry on arterial hypertension, networking 23 community hospital-based hypertension clinics, and 60 general practitioners to the Hypertension Research Center of Federico II University Hospital in Naples,8 from the Campania district in Southern Italy (ClinicalTrials.gov Identifier: NCT02211365). As previously reported in detail,9,10 subjects recruited by participating hospitals or general practitioners are referred for CV imaging to the Hypertension Research Center. The registry currently includes over 15,000 subjects. For the present analysis 7336 hypertensive patients without prevalent coronary or cerebrovascular disease, valvular heart disease, with ejection fraction ≥50%, no more than stage III chronic kidney disease (based on estimation of glomerular filtration rate (GFR) by Chronic Kidney Disease Epidemiology Collaboration formula.)11 and with baseline echocardiogram and carotid ultrasound were selected (Figure 1). Figure 1. Open in new tabDownload slide Selection of the study population. CV: cardiovascular; CKD: Chronic Kidney Disease; EF: ejection fraction. The database generation of the CSN was approved by the Federico II University Hospital Ethic Committee. All participants signed written informed consent for the possibility of using the data for scientific purposes. CV risk factor assessment Prevalent CV disease was defined at the first examination in the outpatient clinic and included previous myocardial infarction (MI), angina pectoris, coronary revascularization procedures, stroke, transitory ischaemic attack, or congestive heart failure.12 Office systolic and diastolic blood pressures (SBP and DBP) were measured in triplets after at least five minutes of rest in the sitting position, following current guidelines.4 The average of the two last measurements was taken as the clinical blood pressure (BP). In the CSN the measurement of BP has been always attended by a physician, using a regularly calibrated aneroid sphygmomanometer. Follow-up (FU) BP was considered controlled when the average clinic BP values during FU < 140/90 mm Hg.8 High PP was classified when PP ≥ 60 mm Hg.4 High systolic BP was ≥140 mm Hg. Obesity was defined as a body mass index (BMI) ≥30 kg/m2.13 Fasting glucose and lipid profile were measured by standard methods. Diabetes was defined as history of diabetes, use of any anti-diabetic medication or presence of a fasting blood glucose ≥126 mg/dl.14 Diabetes was defined controlled if the average value of all glycaemic measurements during FU were <126 mg/dl, and uncontrolled if mean value of all glycaemic measurements during FU was ≥126 mg/dl. Echocardiography Echocardiography was performed using commercially available phased-array machines following a standardised protocol. All final readings were performed off-line by a single highly experienced reader and supervised by a senior faculty member. Quantitative echocardiography was performed following the joint European Association of Echocardiography and American Society of Echocardiography recommendations.15 LVH was identified by prognostically validated sex-specific cut-off values for LV mass/height, >47 g/m2.7 in women and >50 g/m2.7 in men, respectively.16 Consistent with our approach to normalization for body size, LV end-diastolic dimension was normalised by height. Relative wall thickness was calculated as the ratio between posterior wall thickness and LV internal radius at end-diastole and considered increased if ≥0.43.15 LV volumes were estimated by the z-derived method.17,18 Left atrial (LA) volume (eLAV) was estimated from LA antero-posterior diameter in cm (LAd), measured in the longitudinal parasternal window, using a validated non linear equation: eLAV=2.323×LAd2.071 and normalised by height in m2 (eLAVi).19 Carotid ultrasonography Carotid ultrasonography was performed using a commercially available ultrasound scanner equipped with a 7.5-MHz high-resolution transducer with an axial resolution of 0.1 mm. B-mode ultrasonography was performed with subjects in the supine position with the head turned away from the sonographer and the neck extended in mild rotation. Images were recorded on super video home system tapes for off-line analysis. The standardised comprehensive scanning and reading protocol has been previously published.20 The maximal arterial intima media thickness (IMT) was estimated offline in up to 12 arterial walls, including the right and the left, near and far distal common carotid (1 cm), bifurcation and proximal internal carotid artery, according to European Society of Hypertension/European Society of Cardiology guidelines.4 The average and the maximum IMT was reported in the individual patient. Carotid plaque was defined as IMT ≥ 1.5 mm in accordance with previous studies.21 Endpoints Major adverse cardiovascular end-points (MACEs) at their first presentation were considered fatal or nonfatal MI or stroke, and sudden death. Secondary analyses were also performed using composite major and minor end-points, also including first presentation of heart failure requiring hospitalization, incident coronary revascularization, angina, transient ischaemic attack, carotid artery stenting and atrial fibrillation. Both prevalent and incident CV and cerebrovascular events were adjudicated by the Committee for Event Adjudication in the Hypertension Research Center and were based on patient history, contact with the reference general practitioner, and review of hospital medical records.22 Statistical analysis Statistical analysis was performed using IBM SPSS 23 (IBM Corporation, Armonk, New York, USA). Data are presented as mean ± one standard deviation for continuous variables and as percentages for categorical variables. Patients were categorised into two groups according to the absence (normal pulse pressure (NPP); n = 4980) or the presence of high PP (HPP; n = 2356): Analysis of variance was used to compare continuous variables. The χ2 distribution was used to compare categorical variables. As previously reported,22,23 to account for antihypertensive therapy during FU, single classes of medications were dichotomised according to their overall use during the individual FU, based on the frequency of prescriptions at the control visits during FU. All medications prescribed for more than 50% of control visits in an individual patient during FU were considered as covariates in the Cox analyses.22,23 Kaplan–Meier survival curves were used to compare outcome of HPP vs NPP for MACEs. The outcome was also analysed in relation to LVH and carotid plaque in combination with high PP in Cox regression analyses for either categories of PP or PP as continuous variable. Hazard ratios (HRs) and 95% confidence intervals (CIs) were reported. In a first Cox model, the condition of high PP was adjusted for age and sex; in a second model, presence of diabetes, LVH and carotid plaque were forced. A third model also included all classes of antihypertensive medications used during FU. Finally Model 4 was additionally adjusted for mean arterial pressure ((2 × DBP + SBP)/3). The same analysis was done for PP as a continuous variable. All variables used in the Cox models were analysed for multicollinearity, by computing linear variance inflation factor (VIF). VIF was always less than 1.5. A two-tailed p-value < 0.05 was considered statistically significant in all analyses. Results HPP patients were older, more likely to be women and diabetic, and exhibited higher SBP and lower GFR, while receiving more of each class of antihypertensive medications than NPP (all p < 0.0001, Table 1). No difference was observed in lipid profile. Table 1. General characteristic of the study population. Variable . HPP (n = 2356) . NPP (n = 4980) . p . Age (years) 57 ± 12 52 ± 11 0.0001 Women (%) 50 39 0.0001 Systolic blood pressure (mm Hg) 159.6 ± 17 135 ± 13 0.0001 Diastolic blood pressure (mm Hg) 88.5 ± 12.5 88.9 ± 10.4 0.221 Mean blood pressure (mm Hg) 112.1 ± 13.1 105 ± 11.1 0.0001 Heart rate (bpm) 74.7 ± 11.1 74.2 ± 11 0.104 Obesity (%) 25 26 0.256 Diabetes (%) 14 8 0.0001 Estimated glomerular filtration rate (ml/min/1.73m2) 78.7 ± 15.8 82.1 ± 14.9 0.0001 Total serum cholesterol (mg/dl) 208 ± 39.6 206.4 ± 38.9 0.111 Serum HDL cholesterol (mg/dl) 51.2 ± 13.4 50.5 ± 12.9 0.075 Serum non-HDL cholesterol (mg/dl) 129 ± 36.2 128.4 ± 35.6 0.444 Serum triglycerides (mg/dl) 136 ± 78.4 135.7 ± 75.6 0.870 Number of antihypertensive drugs during FU 1.8 ± 1.1 1.5 ± 1.0 0.0001 Anti-RAS during FU (%) 82 77 0.0001 Beta-blockers during FU (%) 29 24 0.0001 Calcium-blockers during FU (%) 33 22 0.0001 Diuretics during FU (%) 49 39 0.0001 Variable . HPP (n = 2356) . NPP (n = 4980) . p . Age (years) 57 ± 12 52 ± 11 0.0001 Women (%) 50 39 0.0001 Systolic blood pressure (mm Hg) 159.6 ± 17 135 ± 13 0.0001 Diastolic blood pressure (mm Hg) 88.5 ± 12.5 88.9 ± 10.4 0.221 Mean blood pressure (mm Hg) 112.1 ± 13.1 105 ± 11.1 0.0001 Heart rate (bpm) 74.7 ± 11.1 74.2 ± 11 0.104 Obesity (%) 25 26 0.256 Diabetes (%) 14 8 0.0001 Estimated glomerular filtration rate (ml/min/1.73m2) 78.7 ± 15.8 82.1 ± 14.9 0.0001 Total serum cholesterol (mg/dl) 208 ± 39.6 206.4 ± 38.9 0.111 Serum HDL cholesterol (mg/dl) 51.2 ± 13.4 50.5 ± 12.9 0.075 Serum non-HDL cholesterol (mg/dl) 129 ± 36.2 128.4 ± 35.6 0.444 Serum triglycerides (mg/dl) 136 ± 78.4 135.7 ± 75.6 0.870 Number of antihypertensive drugs during FU 1.8 ± 1.1 1.5 ± 1.0 0.0001 Anti-RAS during FU (%) 82 77 0.0001 Beta-blockers during FU (%) 29 24 0.0001 Calcium-blockers during FU (%) 33 22 0.0001 Diuretics during FU (%) 49 39 0.0001 GFR: glomerular filtration rate; FU: follow-up; HDL: high density lipoprotein; RAS: renin–angiotensin system. Open in new tab Table 1. General characteristic of the study population. Variable . HPP (n = 2356) . NPP (n = 4980) . p . Age (years) 57 ± 12 52 ± 11 0.0001 Women (%) 50 39 0.0001 Systolic blood pressure (mm Hg) 159.6 ± 17 135 ± 13 0.0001 Diastolic blood pressure (mm Hg) 88.5 ± 12.5 88.9 ± 10.4 0.221 Mean blood pressure (mm Hg) 112.1 ± 13.1 105 ± 11.1 0.0001 Heart rate (bpm) 74.7 ± 11.1 74.2 ± 11 0.104 Obesity (%) 25 26 0.256 Diabetes (%) 14 8 0.0001 Estimated glomerular filtration rate (ml/min/1.73m2) 78.7 ± 15.8 82.1 ± 14.9 0.0001 Total serum cholesterol (mg/dl) 208 ± 39.6 206.4 ± 38.9 0.111 Serum HDL cholesterol (mg/dl) 51.2 ± 13.4 50.5 ± 12.9 0.075 Serum non-HDL cholesterol (mg/dl) 129 ± 36.2 128.4 ± 35.6 0.444 Serum triglycerides (mg/dl) 136 ± 78.4 135.7 ± 75.6 0.870 Number of antihypertensive drugs during FU 1.8 ± 1.1 1.5 ± 1.0 0.0001 Anti-RAS during FU (%) 82 77 0.0001 Beta-blockers during FU (%) 29 24 0.0001 Calcium-blockers during FU (%) 33 22 0.0001 Diuretics during FU (%) 49 39 0.0001 Variable . HPP (n = 2356) . NPP (n = 4980) . p . Age (years) 57 ± 12 52 ± 11 0.0001 Women (%) 50 39 0.0001 Systolic blood pressure (mm Hg) 159.6 ± 17 135 ± 13 0.0001 Diastolic blood pressure (mm Hg) 88.5 ± 12.5 88.9 ± 10.4 0.221 Mean blood pressure (mm Hg) 112.1 ± 13.1 105 ± 11.1 0.0001 Heart rate (bpm) 74.7 ± 11.1 74.2 ± 11 0.104 Obesity (%) 25 26 0.256 Diabetes (%) 14 8 0.0001 Estimated glomerular filtration rate (ml/min/1.73m2) 78.7 ± 15.8 82.1 ± 14.9 0.0001 Total serum cholesterol (mg/dl) 208 ± 39.6 206.4 ± 38.9 0.111 Serum HDL cholesterol (mg/dl) 51.2 ± 13.4 50.5 ± 12.9 0.075 Serum non-HDL cholesterol (mg/dl) 129 ± 36.2 128.4 ± 35.6 0.444 Serum triglycerides (mg/dl) 136 ± 78.4 135.7 ± 75.6 0.870 Number of antihypertensive drugs during FU 1.8 ± 1.1 1.5 ± 1.0 0.0001 Anti-RAS during FU (%) 82 77 0.0001 Beta-blockers during FU (%) 29 24 0.0001 Calcium-blockers during FU (%) 33 22 0.0001 Diuretics during FU (%) 49 39 0.0001 GFR: glomerular filtration rate; FU: follow-up; HDL: high density lipoprotein; RAS: renin–angiotensin system. Open in new tab Table 2 shows that HPP patients exhibited greater LV chamber dimension and stroke index, with greater LV mass and relative wall thickness, greater prevalence of LVH, concentric geometry, and carotid plaque than NPP patients (all p < 0.0001). Table 2. Target organ damage among study population. Variable . HPP (n = 2356) . NPP (n = 4980) . p . LV end-diastolic dimension (cm/m) 3.00 ± 0.19 2.95 ± 0.18 0.0001 Stroke index (ml/beat × m2.04) 26.35 ± 3.5 25.61 ± 3.3 0.0001 Relative wall thickness 0.39 ± 0.04 0.38 ± 0.04 0.0001 LV mass index (g/m2.7) 49.4 ± 9.9 45.7 ± 8.4 0.0001 Intima media thickness (mm) 1.8 ± 0.8 1.5 ± 0.7 0.0001 Concentric LV geometry (%) 12 9 0.0001 LV hypertrophy (%) 49 31 0.0001 Carotid plaque (%) 57 39 0.0001 Variable . HPP (n = 2356) . NPP (n = 4980) . p . LV end-diastolic dimension (cm/m) 3.00 ± 0.19 2.95 ± 0.18 0.0001 Stroke index (ml/beat × m2.04) 26.35 ± 3.5 25.61 ± 3.3 0.0001 Relative wall thickness 0.39 ± 0.04 0.38 ± 0.04 0.0001 LV mass index (g/m2.7) 49.4 ± 9.9 45.7 ± 8.4 0.0001 Intima media thickness (mm) 1.8 ± 0.8 1.5 ± 0.7 0.0001 Concentric LV geometry (%) 12 9 0.0001 LV hypertrophy (%) 49 31 0.0001 Carotid plaque (%) 57 39 0.0001 HPP: high pulse pressure; LV: left ventricular; NPP: normal pulse pressure. Open in new tab Table 2. Target organ damage among study population. Variable . HPP (n = 2356) . NPP (n = 4980) . p . LV end-diastolic dimension (cm/m) 3.00 ± 0.19 2.95 ± 0.18 0.0001 Stroke index (ml/beat × m2.04) 26.35 ± 3.5 25.61 ± 3.3 0.0001 Relative wall thickness 0.39 ± 0.04 0.38 ± 0.04 0.0001 LV mass index (g/m2.7) 49.4 ± 9.9 45.7 ± 8.4 0.0001 Intima media thickness (mm) 1.8 ± 0.8 1.5 ± 0.7 0.0001 Concentric LV geometry (%) 12 9 0.0001 LV hypertrophy (%) 49 31 0.0001 Carotid plaque (%) 57 39 0.0001 Variable . HPP (n = 2356) . NPP (n = 4980) . p . LV end-diastolic dimension (cm/m) 3.00 ± 0.19 2.95 ± 0.18 0.0001 Stroke index (ml/beat × m2.04) 26.35 ± 3.5 25.61 ± 3.3 0.0001 Relative wall thickness 0.39 ± 0.04 0.38 ± 0.04 0.0001 LV mass index (g/m2.7) 49.4 ± 9.9 45.7 ± 8.4 0.0001 Intima media thickness (mm) 1.8 ± 0.8 1.5 ± 0.7 0.0001 Concentric LV geometry (%) 12 9 0.0001 LV hypertrophy (%) 49 31 0.0001 Carotid plaque (%) 57 39 0.0001 HPP: high pulse pressure; LV: left ventricular; NPP: normal pulse pressure. Open in new tab Incident MACEs During the median FU of 41 months (interquartile range = 15–88 months), 157 incident MACEs occurred. Consistent with the risk profile displayed in Tables 1 and 2, HPP patients exhibited 86% greater risk of MACEs than NPP patients (HR = 1.86; 95% CI: 1.36–2.56, p < 0.0001, Figure 2). Figure 2. Open in new tabDownload slide Kaplan–Meier plot of cumulative hazard of major adverse cardiovascular end-points (MACEs) in patients with high vs normal pulse pressure (PP). In sequential Cox regression models, the hazard of HPP was slightly reduced after adjusting for age and sex (Table 3, Model 2). HPP remained a significant predictor of MACEs, also when diabetes, LVH, carotid plaque and antihypertensive therapy were added to the Cox model (Table 3, Model 3). Further control for classes of medicines did not alter the impact of HPP, but revealed that anti-renin–angiotensin system (RAS) medicines were less likely to be prescribed in patients with incident MACEs. Finally, consideration of mean BP did not alter the risk profile (Table 3, Model 4). Even when including control of diabetes in Model 4 displayed in Table 3 the effect of high PP did not change (HR = 1.58, 95% CI 1.12–2.23, p = 0.01). With further adjustment for the presence of obesity and stage III CKD (found in 671 participants), high PP maintained an HR of 1.57, (95% CI 1.11–2.21, p = 0.01). Model 4 in Table 3 has been further adjusted for haemoglobin level, estimated GFR, LA volume index and stroke volume index with high PP maintaining a significant prognostic impact (HR 1.55, 95% CI 1.10–2.20, p = 0.013). To verify whether the introduction of a classical cut-off value of high baseline systolic BP (>140 mm Hg) could influence the prognostic impact of high PP we have run a new model changing mean BP with high systolic BP. As shown in Table 4 the prognostic impact of high PP was not influenced by the introduction of high systolic BP in the model (HR 1.69, 95% CI 1.15–2.49, p = 0.008). Table 3. Sequential models of proportional hazard analysis for major adverse cardiovascular end-point (MACE) adjusting for group of covariates for high pulse pressure (PP). Significant predictors are highlighted in bold. Predictors . Model 1 . Model 2 . Model 3 . Model 4 . Sig. . HR . 95.0% CI . Sig. . HR . 95.0% CI . Sig. . HR . 95.0% CI . Sig. . HR . 95.0% CI . Age (years) 0.0001 1.05 1.03–1.06 0.0001 1.03 1.01–1.05 0.0001 1.03 1.02–1.05 0.0001 1.04 1.02–1.05 Male sex 0.001 1.76 1.25–2.48 0.002 1.72 1.22–2.42 0.002 1.75 1.24–2.48 0.002 1.75 1.23–2.47 High PP (≥60 mm Hg) 0.001 1.72 1.24–2.39 0.007 1.58 1.14–2.20 0.004 1.63 1.16–2.27 0.01 1.57 1.12–2.22 Diabetes (n/y) 0.34 1.59 1.04–2.46 0.019 1.68 1.09–2.59 0.017 1.69 1.01–2.61 LV hypertrophy (n/y) 0.031 1.43 1.03–1.99 0.016 1.51 1.08–2.11 0.019 1.50 1.07–2.1 Carotid plaque (n/y) 0.009 1.59 1.12–2.26 0.004 1.68 1.19–2.39 0.004 1.68 1.18–2.39 Anti–RAS (n/y) 0.0001 0.36 0.24–0.53 0.0001 0.35 0.24–0.53 Beta blockers (n/y) 0.51 1.13 0.79–1.61 0.536 1.12 0.78–1.60 Calcium-channel blockers (n/y) 0.83 0.96 0.68–1.36 0.782 0.95 0.67–1.35 Diuretics (n/y) 0.82 0.96 0.68–1.36 0.794 0.96 0.68–1.35 Mean BP (mm Hg) 0.475 1.01 0.99–1.02 Predictors . Model 1 . Model 2 . Model 3 . Model 4 . Sig. . HR . 95.0% CI . Sig. . HR . 95.0% CI . Sig. . HR . 95.0% CI . Sig. . HR . 95.0% CI . Age (years) 0.0001 1.05 1.03–1.06 0.0001 1.03 1.01–1.05 0.0001 1.03 1.02–1.05 0.0001 1.04 1.02–1.05 Male sex 0.001 1.76 1.25–2.48 0.002 1.72 1.22–2.42 0.002 1.75 1.24–2.48 0.002 1.75 1.23–2.47 High PP (≥60 mm Hg) 0.001 1.72 1.24–2.39 0.007 1.58 1.14–2.20 0.004 1.63 1.16–2.27 0.01 1.57 1.12–2.22 Diabetes (n/y) 0.34 1.59 1.04–2.46 0.019 1.68 1.09–2.59 0.017 1.69 1.01–2.61 LV hypertrophy (n/y) 0.031 1.43 1.03–1.99 0.016 1.51 1.08–2.11 0.019 1.50 1.07–2.1 Carotid plaque (n/y) 0.009 1.59 1.12–2.26 0.004 1.68 1.19–2.39 0.004 1.68 1.18–2.39 Anti–RAS (n/y) 0.0001 0.36 0.24–0.53 0.0001 0.35 0.24–0.53 Beta blockers (n/y) 0.51 1.13 0.79–1.61 0.536 1.12 0.78–1.60 Calcium-channel blockers (n/y) 0.83 0.96 0.68–1.36 0.782 0.95 0.67–1.35 Diuretics (n/y) 0.82 0.96 0.68–1.36 0.794 0.96 0.68–1.35 Mean BP (mm Hg) 0.475 1.01 0.99–1.02 BP: blood pressure; CI: confidence interval; HR: hazard ratio; LV: left ventricular; PP: pulse pressure; RAS: renin–angiotensin system. Open in new tab Table 3. Sequential models of proportional hazard analysis for major adverse cardiovascular end-point (MACE) adjusting for group of covariates for high pulse pressure (PP). Significant predictors are highlighted in bold. Predictors . Model 1 . Model 2 . Model 3 . Model 4 . Sig. . HR . 95.0% CI . Sig. . HR . 95.0% CI . Sig. . HR . 95.0% CI . Sig. . HR . 95.0% CI . Age (years) 0.0001 1.05 1.03–1.06 0.0001 1.03 1.01–1.05 0.0001 1.03 1.02–1.05 0.0001 1.04 1.02–1.05 Male sex 0.001 1.76 1.25–2.48 0.002 1.72 1.22–2.42 0.002 1.75 1.24–2.48 0.002 1.75 1.23–2.47 High PP (≥60 mm Hg) 0.001 1.72 1.24–2.39 0.007 1.58 1.14–2.20 0.004 1.63 1.16–2.27 0.01 1.57 1.12–2.22 Diabetes (n/y) 0.34 1.59 1.04–2.46 0.019 1.68 1.09–2.59 0.017 1.69 1.01–2.61 LV hypertrophy (n/y) 0.031 1.43 1.03–1.99 0.016 1.51 1.08–2.11 0.019 1.50 1.07–2.1 Carotid plaque (n/y) 0.009 1.59 1.12–2.26 0.004 1.68 1.19–2.39 0.004 1.68 1.18–2.39 Anti–RAS (n/y) 0.0001 0.36 0.24–0.53 0.0001 0.35 0.24–0.53 Beta blockers (n/y) 0.51 1.13 0.79–1.61 0.536 1.12 0.78–1.60 Calcium-channel blockers (n/y) 0.83 0.96 0.68–1.36 0.782 0.95 0.67–1.35 Diuretics (n/y) 0.82 0.96 0.68–1.36 0.794 0.96 0.68–1.35 Mean BP (mm Hg) 0.475 1.01 0.99–1.02 Predictors . Model 1 . Model 2 . Model 3 . Model 4 . Sig. . HR . 95.0% CI . Sig. . HR . 95.0% CI . Sig. . HR . 95.0% CI . Sig. . HR . 95.0% CI . Age (years) 0.0001 1.05 1.03–1.06 0.0001 1.03 1.01–1.05 0.0001 1.03 1.02–1.05 0.0001 1.04 1.02–1.05 Male sex 0.001 1.76 1.25–2.48 0.002 1.72 1.22–2.42 0.002 1.75 1.24–2.48 0.002 1.75 1.23–2.47 High PP (≥60 mm Hg) 0.001 1.72 1.24–2.39 0.007 1.58 1.14–2.20 0.004 1.63 1.16–2.27 0.01 1.57 1.12–2.22 Diabetes (n/y) 0.34 1.59 1.04–2.46 0.019 1.68 1.09–2.59 0.017 1.69 1.01–2.61 LV hypertrophy (n/y) 0.031 1.43 1.03–1.99 0.016 1.51 1.08–2.11 0.019 1.50 1.07–2.1 Carotid plaque (n/y) 0.009 1.59 1.12–2.26 0.004 1.68 1.19–2.39 0.004 1.68 1.18–2.39 Anti–RAS (n/y) 0.0001 0.36 0.24–0.53 0.0001 0.35 0.24–0.53 Beta blockers (n/y) 0.51 1.13 0.79–1.61 0.536 1.12 0.78–1.60 Calcium-channel blockers (n/y) 0.83 0.96 0.68–1.36 0.782 0.95 0.67–1.35 Diuretics (n/y) 0.82 0.96 0.68–1.36 0.794 0.96 0.68–1.35 Mean BP (mm Hg) 0.475 1.01 0.99–1.02 BP: blood pressure; CI: confidence interval; HR: hazard ratio; LV: left ventricular; PP: pulse pressure; RAS: renin–angiotensin system. Open in new tab Table 4. Cox model of proportional hazard analysis for major adverse cardiovascular end-point (MACE) adjusting for group of covariates for high pulse pressure (PP) including high systolic blood pressure (BP). Significant predictors are highlighted in bold. Predictors . Sig. . HR . 95.0% CI . Age (years) 0.000 1.034 1.016 −1.052 Male sex 0.002 1.753 1.238 −2.482 High PP (≥60 mm Hg) 0.008 1.689 1.147 −2.487 Diabetes (n/y) 0.019 1.682 1.091 −2.594 LV hypertrophy (n/y) 0.015 1.517 1.085 −2.122 Carotid plaque (n/y) 0.004 1.684 1.184 −2.393 Anti-RAS (n/y) 0.000 0.356 0.239 −0.530 Beta-blockers (n/y) 0.510 1.128 0.788 −1.613 Calcium-channel blockers (n/y) 0.839 0.965 0.682 −1.365 Diuretics (n/y) 0.828 0.963 0.683 −1.357 High systolic BP (≥140 mm Hg) 0.700 0.925 0.623 −1.374 Predictors . Sig. . HR . 95.0% CI . Age (years) 0.000 1.034 1.016 −1.052 Male sex 0.002 1.753 1.238 −2.482 High PP (≥60 mm Hg) 0.008 1.689 1.147 −2.487 Diabetes (n/y) 0.019 1.682 1.091 −2.594 LV hypertrophy (n/y) 0.015 1.517 1.085 −2.122 Carotid plaque (n/y) 0.004 1.684 1.184 −2.393 Anti-RAS (n/y) 0.000 0.356 0.239 −0.530 Beta-blockers (n/y) 0.510 1.128 0.788 −1.613 Calcium-channel blockers (n/y) 0.839 0.965 0.682 −1.365 Diuretics (n/y) 0.828 0.963 0.683 −1.357 High systolic BP (≥140 mm Hg) 0.700 0.925 0.623 −1.374 CI: confidence interval; HR: hazard ratio; LV: left ventricular; RAS: renin–angiotensin system. Open in new tab Table 4. Cox model of proportional hazard analysis for major adverse cardiovascular end-point (MACE) adjusting for group of covariates for high pulse pressure (PP) including high systolic blood pressure (BP). Significant predictors are highlighted in bold. Predictors . Sig. . HR . 95.0% CI . Age (years) 0.000 1.034 1.016 −1.052 Male sex 0.002 1.753 1.238 −2.482 High PP (≥60 mm Hg) 0.008 1.689 1.147 −2.487 Diabetes (n/y) 0.019 1.682 1.091 −2.594 LV hypertrophy (n/y) 0.015 1.517 1.085 −2.122 Carotid plaque (n/y) 0.004 1.684 1.184 −2.393 Anti-RAS (n/y) 0.000 0.356 0.239 −0.530 Beta-blockers (n/y) 0.510 1.128 0.788 −1.613 Calcium-channel blockers (n/y) 0.839 0.965 0.682 −1.365 Diuretics (n/y) 0.828 0.963 0.683 −1.357 High systolic BP (≥140 mm Hg) 0.700 0.925 0.623 −1.374 Predictors . Sig. . HR . 95.0% CI . Age (years) 0.000 1.034 1.016 −1.052 Male sex 0.002 1.753 1.238 −2.482 High PP (≥60 mm Hg) 0.008 1.689 1.147 −2.487 Diabetes (n/y) 0.019 1.682 1.091 −2.594 LV hypertrophy (n/y) 0.015 1.517 1.085 −2.122 Carotid plaque (n/y) 0.004 1.684 1.184 −2.393 Anti-RAS (n/y) 0.000 0.356 0.239 −0.530 Beta-blockers (n/y) 0.510 1.128 0.788 −1.613 Calcium-channel blockers (n/y) 0.839 0.965 0.682 −1.365 Diuretics (n/y) 0.828 0.963 0.683 −1.357 High systolic BP (≥140 mm Hg) 0.700 0.925 0.623 −1.374 CI: confidence interval; HR: hazard ratio; LV: left ventricular; RAS: renin–angiotensin system. Open in new tab To verify whether the introduction of high PP in a traditional TOD-based model influences the hazard of CV events due to structural TOD, another Cox analysis was run forcing the condition of high PP into a model including age, sex, LVH and plaque. The age and sex-adjusted hazard of LVH and carotid plaque were respectively 1.54 (1.11–2.13) and 1.66 (1.17–2.35), decreasing to the value of 1.42 and 1.63 when high PP coexisted in the model (all p < 0.0001). Analysis of PP as a continuous variable yielded similar results (HR 1.15/10 mm Hg (95% CI 1.05–1.25), p = 0.003) independently of confounders. A similar model as displayed in Table 3, Model 3, was run changing PP with its two components, SBP and DBP. In this model, both SBP and DBP were independent predictors of CV events, but in opposite directions, because SBP exhibited a direct relation with probability of events (HR = 1.08/5 mm Hg, 1.05–1.13, p < 0.001), whereas the relation of DBP was negative (HR = 0.89/5 mm Hg, 0.80–0.99, p < 0.03). The fourth Cox regression in Table 3 was also run in a subgroup of patients with controlled BP during FU (3910 patients with 91 MACEs) and high PP retained a tendency for increased CV risk (HR 1.60; 95 CI 1.00–2.55, p = 0.05). Incident composite MACEs and minor end-points During FU, 313 incident composite MACEs and minor end-points occurred. HPP patients exhibited 66% greater risk of composite end-points than NPP patients (odds ratio (OR) = 1.66; 95% CI: 1.32–2.01, p < 0.0001, Figure 3). Figure 3. Open in new tabDownload slide Kaplan–Meier plot of cumulative hazard for composite major adverse cardiovascular end-points (MACEs) and minor events in patients with high vs normal pulse pressure (PP). Similar to analysis with MACEs, in Cox regression analysis, HPP patients maintained a 33% increased hazard, compared to NPP patients (HR 1.33 (95% CI 1.04–1.70), p = 0.024), independently of significant effect of old age, male gender, LVH, carotid plaque and less anti-RAS therapy, without significant effect for mean BP and the other classes of medicines (not shown). Discussion This analysis is the first direct demonstration that high PP is a functional equivalent of a structural TOD, also independent of other commonly measured TODs. This finding is particularly relevant, because it was revealed in a context of primary prevention in a real-world context. PP is the measure of the combined LV ability to expel blood, depending on the capacitance and distensibility of the conduit arterial system.1 Thus, PP represents the pulsatile component of circulatory physiology, a consequence of the balance between systolic and diastolic components which sustain the peripheral circulation. Mean BP estimates the continuous peripheral flow, which is the physiological consequence of the combination of systolic flow with the diastolic arterial run-off at the aortic valve closure. With the progressive stiffening of the conduit arteries, the ability of arterial run-off is lost and consequently the diastolic component of the peripheral flow decreases; if the final goal is to preserve peripheral continuous flow (i.e. mean BP), any increase in stiffening of conduit arteries needs a corresponding increase in the systolic component of peripheral flow (i.e. SBP), which participates to widening of pulse pressure.24 Model 4 of Table 3, with the strong prognostic effect of high PP and the evidence that mean BP does not independently contribute to outlining risk, indicates that the way to maintain peripheral blood flow is substantially important for cardiovascular health. Combining information of both components of BP (SBP + DBP) has been demonstrated to better predict outcome in several studies, especially in the middle-aged population.25,26 The widening of PP due to the diverging of SBP and DBP with aging carries an adverse CV prognosis that includes effects of both higher SBP and lower DBP, which may be more easily identified using PP as a prognostic maker. In addition to BP components, aging is strongly associated with HPP, mainly due to age-related alterations in collagen to elastin ratio and disruption of elastic fibres27 in the conductance arteries. In our outpatient setting of adult hypertensive patients, HPP was also associated with greater prevalence of TOD, suggesting that HPP is another sign of established CV damage in the setting of arterial hypertension.3 This association has been also confirmed on the large cohort of patients of the Multi-Ethnic Study of Atherosclerosis (MESA), where increased PP was correlated with markers of subclinical CV disease and this association was stronger with advancing age.28 Recent data from the Hypertension Genetic Network (HyperGEN) study highlighted the importance of PP for the development of subclinical cardiac target organ damage, in both hypertensive and non-hypertensive patients.29 Other studies have demonstrated that both subclinical TOD and CV prognosis were related to increased PP, suggesting a specific cut-off (>60 mm Hg) in study populations older than 60 years.30,31 The prognostic impact of PP on CV disease has been recently demonstrated in a population-based study with high prevalence of obesity and hypertension.6 A pulse pressure ≥70 mm Hg increased risk of CV events, even after adjustment for multiple confounders such as age, sex, smoking, hypercholesterolemia and other confounders. Comparing with our study, the Reach Registry included also patients with prevalent CV diseases who were not included in our study, focused on primary prevention. Another large longitudinal cohort study indicated that PP higher than 65 mm Hg was an independent risk factor for acute coronary heart disease even after adjustment for SBP.32 Particularly interesting is the observation that when the classical cut-off value of high systolic BP coexists in the same Cox regression model with high PP the risk associated with systolic BP ≥ 140 mm Hg is completely obscured by risk represented by a PP ≥ 60 mm Hg. No study, however, verified whether the prognostic impact of PP was independent of other markers of TOD, since no ultrasound data were reported. Taking advantages of our large registry of hypertensive outpatients, we could demonstrate that increased PP is a potent predictor of CV events, and this prognostic power was maintained not only in addition to a large number of confounders and antihypertensive therapy, but also independently of common markers of TOD, such as LVH and carotid plaque. This outcome effect independent of structural TOD could be documented either with MACEs or with composite MACEs and minor events. It should be underlined that, in the exploratory analysis, patients with high PP were more likely to be women and more treated with anti-RAS therapy during FU. However, in Cox regression, those who suffered with MACEs were more frequently men and had a reduced probability of having been treated with anti-RAS therapy. The modification of HRs of LVH and carotid plaque when there is high PP in the Cox model suggests that all three factors explore different aspect of the pre-clinical CV disease pointing towards pathophysiologic links. Exposure to risk factors represents the initial stimulus to induce biological changes leading to structural abnormalities in the arterial system and at cardiac level.33 Arteriosclerosis and consequent stiffening of conduit arteries reduce the arterial run-off due to elastic recoil, with consequences on maintenance of DBP, a process that is also proposed to change the initial presentation of hypertension and lead to isolated systolic hypertension.34,35 LVH develops as the consequence of haemodynamic modifications and takes the geometry determined by the balance between pressure and volume overload.16,22 PP increases as the consequence of merging reduced arterial compliance with maintained or even increased stroke volume, regulated by the dominant haemodynamic workload and the availability of contractile mass.16,36 Atherosclerosis, manifested by the presence of carotid plaque, is a particularly severe manifestation of the arterial system impairment. Based on the evidence provided by the present analysis, and on the recalled considerations, the link between structural TOD (i.e. LVH and carotid atherosclerosis) and PP is evident in the process of defining a specific risk profile. We propose considering ‘high PP’ in the range of age analysed in the CSN as a ‘functional’ TOD to merge with structural TOD.37 As a functional marker of TOD, increased PP identifies a status in which the vascular ventricular interaction is no longer working properly,38 especially in older age, when the amplification of the pulsatile wave toward peripheral arteries substantially attenuates and, therefore, cuff PP is closer to central PP.39 In a large registry of hypertensive patients free of prevalent CV disease, high PP is a potent marker of TOD, predicting CV events independently of LV hypertrophy and carotid plaque with a cut off value of 60 mm Hg. High PP is a functional marker of TOD paralleling the effect of structural TOD. Limitations The CSN is an observational registry which can be influenced by bias, a limitation that is difficult to eliminate despite the extensive multivariable adjustment that we performed. However, we paid particular attention to minimising both selection and observational bias, by receiving all hypertensive patients seen in our network and applying substantially the same protocol to everyone. It needs to be underlined that observational studies cannot demonstrate any cause–effect relationship, but are useful for the generation of hypotheses to be tested in prospective studies and/or clinical trials. The measurement of PP is a rough method to estimate arterial stiffness. However, there is evidence that what is thought to be the gold standard to measure arterial stiffness, pulse wave velocity, is closely related to pulse pressure especially with advancing age.40 In addition, our findings have direct, easy clinical applicability in daily clinical practice, with the implicit recommendation to doctors to consider PP in addition to SBP and DBP, a practice that is not yet very common. Author contribution CM and GdS contributed to the conception and design of the work. MAL, RI, GC, MVC, GA contributed to the acquisition, analysis and interpretation of data for the work. CM and NDL drafted the manuscript. BT critically revised the manuscript. All gave final approval and agree to be accountable for all aspects of work ensuring integrity and accuracy. Acknowledgment The authors wish to thank all of the nurses for their support in the management of patients. Declaration of conflicting interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. Funding The author(s) received no financial support for the research, authorship, and/or publication of this article. References 1 Dart AM , Kingwell BA. Pulse pressure–a review of mechanisms and clinical relevance . J Am Coll Cardiol 2001 ; 37 : 975 – 984 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Domanski MJ , Mitchell GF, Norman JEet al. Independent prognostic information provided by sphygmomanometrically determined pulse pressure and mean arterial pressure in patients with left ventricular dysfunction . J Am Coll Cardiol 1999 ; 33 : 951 – 958 . Google Scholar Crossref Search ADS PubMed WorldCat 3 de Simone G , Roman MJ, Alderman MHet al. Is high pulse pressure a marker of preclinical cardiovascular disease? Hypertension 2005 ; 45 : 575 – 579 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Mancia G , Fagard R, Narkiewicz Ket al. 2013 ESH/ESC guidelines for the management of arterial hypertension: The Task Force for the Management of Arterial Hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC) . Eur Heart J 2013 ; 34 : 2159 – 2219 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Roman MJ , Devereux RB, Kizer JRet al. High central pulse pressure is independently associated with adverse cardiovascular outcome the strong heart study . J Am Coll Cardiol 2009 ; 54 : 1730 – 1734 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Selvaraj S , Steg PG, Elbez Yet al. Pulse pressure and risk for cardiovascular events in patients with atherothrombosis: From the REACH Registry . J Am Coll Cardiol 2016 ; 67 : 392 – 403 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Yu D , Simmons D. Association between pulse pressure and risk of hospital admissions for cardiovascular events among people with Type 2 diabetes: A population-based case-control study . Diabet Med 2015 ; 32 : 1201 – 1206 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Mancusi C, Izzo R, de Simone G, et al. Determinants of decline of renal function in treated hypertensive patients: The Campania Salute Network. Nephrol Dial Transplant 2018, in press . 9 Izzo R , de Simone G, Trimarco Vet al. Hypertensive target organ damage predicts incident diabetes mellitus . Eur Heart J 2013 ; 34 : 3419 – 3426 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Lønnebakken MT , Izzo R, Mancusi Cet al. Aortic root dimension and arterial stiffness in arterial hypertension: The Campania Salute Network . J Hypertens 2016 ; 34 : 1109 – 1114 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Levey AS , Stevens LA, Schmid CHet al. A new equation to estimate glomerular filtration rate . Ann Intern Med 2009 ; 150 : 604 – 612 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Mancusi C , Losi MA, Izzo Ret al. Effect of diabetes and metabolic syndrome on myocardial mechano-energetic efficiency in hypertensive patients. The Campania Salute Network . J Hum Hypertens 2017 ; 31 : 424 – 424 . Google Scholar Crossref Search ADS PubMed WorldCat 13 de Simone G , Mancusi C, Izzo Ret al. Obesity and hypertensive heart disease: Focus on body composition and sex differences . Diabetol Metab Syndr 2016 ; 8 : 79 – 79 . Google Scholar Crossref Search ADS PubMed WorldCat 14 America Diabetes Association . Diagnosis and classification of diabetes mellitus . Diabetes Care 2007 ; 30 : S42 – S47 . Crossref Search ADS PubMed WorldCat 15 Marwick TH , Gillebert TC, Aurigemma Get al. Recommendations on the use of echocardiography in adult hypertension: A report from the European Association of Cardiovascular Imaging (EACVI) and the American Society of Echocardiography (ASE) . Eur Heart J Cardiovasc Imaging 2015 ; 16 : 577 – 605 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 16 de Simone G , Izzo R, Aurigemma GPet al. Cardiovascular risk in relation to a new classification of hypertensive left ventricular geometric abnormalities . J Hypertens 2015 ; 33 : 745 – 754 . Google Scholar Crossref Search ADS PubMed WorldCat 17 de Simone G , Devereux RB, Ganau Aet al. Estimation of left ventricular chamber and stroke volume by limited M-mode echocardiography and validation by two-dimensional and Doppler echocardiography . Am J Cardiol 1996 ; 78 : 801 – 807 . Google Scholar Crossref Search ADS PubMed WorldCat 18 De Marco M , Gerdts E, Mancusi Cet al. Influence of left ventricular stroke volume on incident heart failure in a population with preserved ejection fraction (from the Strong Heart Study) . Am J Cardiol 2017 ; 119 : 1047 – 1052 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Canciello G , de Simone G, Izzo Ret al. Validation of left atrial volume estimation by left atrial diameter from the parasternal long-axis view . J Am Soc Echocardiogr 2017 ; 30 : 262 – 269 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Losi MA , Izzo R, De Marco Met al. Cardiovascular ultrasound exploration contributes to predict incident atrial fibrillation in arterial hypertension: The Campania Salute Network . Int J Cardiol 2015 ; 199 : 290 – 295 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Mancusi C , Gerdts E, Losi MAet al. Differential effect of obesity on prevalence of cardiac and carotid target organ damage in hypertension (the Campania Salute Network) . Int J Cardiol 2017 ; 244 : 260 – 264 . Google Scholar Crossref Search ADS PubMed WorldCat 22 de Simone G, Izzo R, Losi MA, et al. Depressed myocardial energetic efficiency is associated with increased cardiovascular risk in hypertensive left ventricular hypertrophy. J Hypertens 2016; 34: 1846–53 . 23 Lønnebakken MT, Izzo R, Mancusi C, et al. Left ventricular hypertrophy regression during antihypertensive treatment in an outpatient clinic (the Campania Salute Network). J Am Heart Assoc 2017; 6. doi: 10.1161/JAHA.116.004152 . 24 Gavish B , Izzo JL Jr. Arterial stiffness: Going a step beyond . Am J Hypertens 2016 ; 29 : 1223 – 1233 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Franklin SS , Lopez VA, Wong NDet al. Single versus combined blood pressure components and risk for cardiovascular disease: The Framingham Heart Study . Circulation 2009 ; 119 : 243 – 250 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Domanski M , Mitchell G, Pfeffer Met al. Pulse pressure and cardiovascular disease-related mortality: Follow-up study of the Multiple Risk Factor Intervention Trial (MRFIT) . JAMA 2002 ; 287 : 2677 – 2683 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Franklin SS , Gustin W 4th, Wong NDet al. Hemodynamic patterns of age-related changes in blood pressure. The Framingham Heart Study . Circulation 1997 ; 96 : 308 – 315 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Winston GJ , Palmas W, Lima Jet al. Pulse pressure and subclinical cardiovascular disease in the Multi-Ethnic Study of Atherosclerosis . Am J Hypertens 2013 ; 26 : 636 – 642 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Glasser SP , Krasikov T, Devereux RBet al. Subclinical, hemodynamic, and echocardiographic abnormalities of high pulse pressure in hypertensive and non-hypertensive adults . Am J Cardiovasc Dis 2012 ; 2 : 309 – 317 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 30 Franklin SS , Larson MG, Khan SAet al. Does the relation of blood pressure to coronary heart disease risk change with aging? The Framingham Heart Study . Circulation 2001 ; 103 : 1245 – 1249 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Anderson RJ , Bahn GD, Emanuele NVet al. Blood pressure and pulse pressure effects on renal outcomes in the Veterans Affairs Diabetes Trial (VADT) . Diabetes Care 2014 ; 37 : 2782 – 2788 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Glasser SP , Halberg DL, Sands Cet al. Is pulse pressure an independent risk factor for incident acute coronary heart disease events? The REGARDS study . Am J Hypertens 2014 ; 27 : 555 – 563 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Devereux RB , Alderman MH. Role of preclinical cardiovascular disease in the evolution from risk factor exposure to development of morbid events . Circulation 1993 ; 88 : 1444 – 1455 . Google Scholar Crossref Search ADS PubMed WorldCat 34 Mancusi C , Gerdts E, de Simone Get al. Impact of isolated systolic hypertension on normalization of left ventricular structure during antihypertensive treatment (the LIFE study) . Blood Press 2014 ; 23 : 206 – 212 . Google Scholar Crossref Search ADS PubMed WorldCat 35 Esposito R , Izzo R, Galderisi Met al. Identification of phenotypes at risk of transition from diastolic hypertension to isolated systolic hypertension . J Hum Hypertens 2016 ; 30 : 392 – 396 . Google Scholar Crossref Search ADS PubMed WorldCat 36 Mancusi C , Gerdts E, de Simone Get al. Higher pulse pressure/stroke volume index is associated with impaired outcome in hypertensive patients with left ventricular hypertrophy the LIFE study . Blood Press 2017 ; 26 : 150 – 155 . Google Scholar Crossref Search ADS PubMed WorldCat 37 Tsao CW , Pencina KM, Massaro JMet al. Cross-sectional relations of arterial stiffness, pressure pulsatility, wave reflection, and arterial calcification . Arterioscler Thromb Vasc Biol 2014 ; 34 : 2495 – 2500 . Google Scholar Crossref Search ADS PubMed WorldCat 38 Zieman SJ , Melenovsky V, Kass DA. Mechanisms, pathophysiology, and therapy of arterial stiffness . Arterioscler Thromb Vasc Biol 2005 ; 25 : 932 – 943 . Google Scholar Crossref Search ADS PubMed WorldCat 39 Roman MJ , Devereux RB, Kizer JRet al. Central pressure more strongly relates to vascular disease and outcome than does brachial pressure: The Strong Heart Study . Hypertension 2007 ; 50 : 197 – 203 . Google Scholar Crossref Search ADS PubMed WorldCat 40 Schillaci G , Mannarino MR, Pucci Get al. Age-specific relationship of aortic pulse wave velocity with left ventricular geometry and function in hypertension . Hypertension 2007 ; 49 : 317 – 321 . Google Scholar Crossref Search ADS PubMed WorldCat © The European Society of Cardiology 2018 This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) © The European Society of Cardiology 2018

Abstract Background Although psychological interventions are recommended for the management of coronary heart disease (CHD), there remains considerable uncertainty regarding their effectiveness. Design Systematic review and meta-analysis of randomised controlled trials (RCTs) of psychological interventions for CHD. Methods The Cochrane Central Register of Controlled Trials, MEDLINE, EMBASE, CINAHL and PsycINFO were searched to April 2016. Retrieved papers, systematic reviews and trial registries were hand-searched. We included RCTs with at least 6 months of follow-up, comparing the direct effects of psychological interventions to usual care for patients following myocardial infarction or revascularisation or with a diagnosis of angina pectoris or CHD defined by angiography. Two authors screened titles for inclusion, extracted data and assessed risk of bias. Studies were pooled using random effects meta-analysis and meta-regression was used to explore study-level predictors. Results Thirty-five studies with 10,703 participants (median follow-up 12 months) were included. Psychological interventions led to a reduction in cardiovascular mortality (rfcelative risk 0.79, 95% confidence interval [CI] 0.63 to 0.98), although no effects were observed for total mortality, myocardial infarction or revascularisation. Psychological interventions improved depressive symptoms (standardised mean difference [SMD] –0.27, 95% CI –0.39 to –0.15), anxiety (SMD –0.24, 95% CI –0.38 to –0.09) and stress (SMD –0.56, 95% CI –0.88 to –0.24) compared with controls. Conclusions We found that psychological intervention improved psychological symptoms and reduced cardiac mortality for people with CHD. However, there remains considerable uncertainty regarding the magnitude of these effects and the specific techniques most likely to benefit people with different presentations of CHD. Cardiac morbidity, mortality, depression, anxiety, stress, psychological intervention, systematic review, randomised controlled trial Introduction Coronary heart disease (CHD) is the single leading cause of death globally, accounting for around a third of all deaths.1 This mortality rate is falling and many more people are living with CHD and require support to manage their symptoms and prognosis. Cardiac events or cardiac surgery can be significant and distressing life events; mental health comorbidity is common, greatly exceeding the rates observed within the general population.2,3 Anxiety and depression are also independent risk factors for cardiovascular morbidity and mortality.4,5 Thus, the need to address stress, psychosocial factors (e.g. lack of social support) and other underlying mood disorders is recognised within conventional cardiac care in Australia,6 Europe7,8 and the USA.4 A range of psychological therapies have been employed as part of secondary prevention to improve psychological outcomes (as opposed to facilitating cardiovascular risk factor reduction). Examples include relaxation and stress management, treatments for mood disorders or enhancing disease adjustment and coping strategies. Therapies have been used both in unselected cardiac populations or targeted at cardiac patients with established psychopathologies. In 2011, a Cochrane review9 synthesised 24 trials testing the direct effects of psychological interventions on cardiac and psychological outcomes compared with usual care. This review observed marked variation in the psychological interventions tested across studies. A meta-analysis found no conclusive evidence that psychological interventions had an effect on total mortality and cardiovascular morbidity, although a potential effect on cardiac mortality was observed (five trials, 3893 participants; relative risk [RR] 0.80, 95% confidence interval [CI] 0.64 to 1.00). There was some evidence that psychological interventions improved depressive symptoms (12 trials, 5041 participants; standardised mean difference [SMD] –0.21, 95% CI –0.48 to –0.08) and anxiety (eight trials, 2771 participants; SMD –0.25, 95% CI –0.35 to –0.03), although the 95% CIs were wide and estimates lacked precision. This paper is an update of this Cochrane review, which is needed now due to the publication of a number of relevant new trials, combined with the considerable uncertainties in the evidence regarding the impact of psychological interventions on clinical events, psychological outcomes and health-related quality of life. Methods We conducted this third update of this Cochrane review10 in accordance the Cochrane Handbook11 and reported it following the PRISMA guidance12 (see Supplementary Figure 1 for the PRISMA flow chart). Although the protocol was first published on the Cochrane Database of Systematic Reviews in 2000, this review builds on the substantively revised protocol implemented in the second update.9 Data searches and sources Search terms from the 2011 review9 were updated and CENTRAL (Cochrane Central Register of Controlled Trials), MEDLINE (Ovid), EMBASE (Ovid), PsychINFO (Ovid) and CINAHL (EBSCO) were searched up to April 2016. We searched the World Health Organization (WHO) International Clinical Trials Registry Platform and the US ClinicalTrials.gov registry for active clinical trials (accessed June 2016). No language limitations were imposed on the searches (Supplementary Methods 1). Study selection We selected randomised controlled trials (RCTs) comparing the direct effects of a psychological intervention compared with a usual care control group for adults with CHD, with or without clinical psychopathology. Participants included those who had experienced a myocardial infarction (MI), a revascularisation procedure (coronary artery bypass grafting [CABG] or percutaneous coronary intervention [PCI]), angina pectoris or angiographically defined CHD. Participants could receive cardiac rehabilitation as long as this was part of usual medical care and offered routinely to both trial arms. Studies where psycho-pharmacology was offered solely or disproportionately to the treatment group in conjunction with the offer of psychological interventions were included. Studies testing psychological interventions in comorbid populations (e.g. patients with depression and either CHD or diabetes) were deemed eligible for inclusion as long as outcome data could be extracted for individuals with CHD. We excluded studies where over 50% of the sample had other cardiac conditions (e.g. heart failure) or had undergone cardiac resynchronisation therapy or received implantable defibrillators. Eligible interventions included those addressing stress or low mood or enhancing coping strategies, either alone or in combination. Studies evaluating interventions based on psychological principles (e.g. motivational interviewing), which were solely directed at improving adherence to other efficacious treatments (e.g. medication adherence or exercise) or the modification of cardiac risk factors (e.g. smoking or diet), were excluded. We only selected studies where the psychological interventions were delivered by health care workers who had been trained in their delivery. Finally, we selected trials reporting outcomes for a minimum of 6 months post-randomisation and reporting at least one of the primary outcomes (reported below). Two reviewers (LA and SHR or CEJ) independently assessed all identified titles/abstracts for possible inclusion, with full reports obtained and assessed for any potentially relevant references. Any disagreements between the reviewers were resolved by discussion. Where necessary, studies were translated into English. Data extraction and management Event rate data were extracted for the dichotomous primary outcomes of total mortality, cardiac mortality and cardiovascular morbidity (non-fatal MI and revascularisation procedures [CABG and PCI]). Means and standard deviations were extracted for the continuous primary outcomes of validated measures assessing symptoms of depression, anxiety or stress. In addition, data were extracted for secondary outcomes regarding other validated measures of psychological function, health-related quality of life (HRQL) and cost-effectiveness. One reviewer (LA) extracted study and participant characteristics, intervention and comparator descriptors and outcomes from included studies using a standardised data extraction form. A second author (SHR or CEJ) checked the extracted data for accuracy and disagreements were resolved by discussion. Outcome data were independently extracted by two reviewers (LA and SHR). Related publications of the same study were assessed for additional data. Authors were contacted, where necessary, to provide additional information. Assessment of risk of bias and overall quality of evidence The Cochrane Collaboration’s core risk of bias items and three further items deemed relevant to this review were assessed, with each study assigned a ‘low’, ‘high’ or ‘unclear’ risk of bias for each item. A detailed description for the three additional criteria (groups balanced at baseline; use of intention-to-treat analysis; and groups receiving comparable treatment except the psychological treatment) can be found elsewhere.10 One reviewer extracted these data and a second reviewer checked the extracted data for accuracy. For each outcome, the overall quality of evidence was assessed by employing the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach to interpret result findings using GRADEpro GDT software.13 Data synthesis and analysis Dichotomous outcomes relating to mortality and cardiovascular morbidity were expressed as risk ratios with 95% CIs. Continuous outcomes relating to psychological outcomes were expressed as SMDs with 95% CIs. For primary outcomes, data were pooled using a conservative random effects model due to the substantial clinical heterogeneity in psychological treatments and study populations identified. Heterogeneity was explored qualitatively and quantitatively (using the I2 statistic and chi-square test of heterogeneity). Small study bias was examined through visual inspection of the funnel plot and the use of Egger tests.14 For secondary outcomes where there were insufficient data or where it was inappropriate to combine studies statistically, a narrative review was presented. Exploratory meta-regression was undertaken to examine potential treatment effect modifiers (Table 1) on the selected outcomes of total mortality, cardiac mortality, depression and anxiety. The explanatory variables were selected a priori following the approach outlined in the 2011 update,9 although we restricted analyses to a smaller group of variables due to concerns over data quality. Given the relatively small ratio of trials to covariates, meta-regression was limited to univariate analysis.15 Table 1. Potential explanatory variables explored in univariate meta-regression. Variable . Levelsa . Targeting of psychological interventions ‘non-selected population (including not reported)’, ‘population with clinically established psychological disorder’ Mode of intervention delivery ‘individual (including not reported)’, ‘group or mix of individual & group’ Family involvement in intervention ‘no (including not reported)’, ‘yes’ Cardiac risk factor education included as part of the intervention ‘no (including not reported)’, ‘yes’ Behaviour change for cardiac risk factors included as part of the intervention ‘no (including not reported)’, ‘yes’ Psychological treatment targets  Depression ‘no (including not reported)’, ‘yes’  Anxiety ‘no (including not reported)’, ‘yes’  Stress management ‘no (including not reported)’, ‘yes’  Type A behaviour ‘no (including not reported)’, ‘yes’ Psychological components  Relaxation ‘no (including not reported)’, ‘yes’  Cognitive techniques ‘no (including not reported)’, ‘yes’  Emotional support and/or client-led discussion ‘no (including not reported)’, ‘yes’  Adjunct pharmacology ‘no (including not reported)’, ‘yes’ Variable . Levelsa . Targeting of psychological interventions ‘non-selected population (including not reported)’, ‘population with clinically established psychological disorder’ Mode of intervention delivery ‘individual (including not reported)’, ‘group or mix of individual & group’ Family involvement in intervention ‘no (including not reported)’, ‘yes’ Cardiac risk factor education included as part of the intervention ‘no (including not reported)’, ‘yes’ Behaviour change for cardiac risk factors included as part of the intervention ‘no (including not reported)’, ‘yes’ Psychological treatment targets  Depression ‘no (including not reported)’, ‘yes’  Anxiety ‘no (including not reported)’, ‘yes’  Stress management ‘no (including not reported)’, ‘yes’  Type A behaviour ‘no (including not reported)’, ‘yes’ Psychological components  Relaxation ‘no (including not reported)’, ‘yes’  Cognitive techniques ‘no (including not reported)’, ‘yes’  Emotional support and/or client-led discussion ‘no (including not reported)’, ‘yes’  Adjunct pharmacology ‘no (including not reported)’, ‘yes’ a First level coded ‘0’ and second level coded ‘1’ in regression models. Open in new tab Table 1. Potential explanatory variables explored in univariate meta-regression. Variable . Levelsa . Targeting of psychological interventions ‘non-selected population (including not reported)’, ‘population with clinically established psychological disorder’ Mode of intervention delivery ‘individual (including not reported)’, ‘group or mix of individual & group’ Family involvement in intervention ‘no (including not reported)’, ‘yes’ Cardiac risk factor education included as part of the intervention ‘no (including not reported)’, ‘yes’ Behaviour change for cardiac risk factors included as part of the intervention ‘no (including not reported)’, ‘yes’ Psychological treatment targets  Depression ‘no (including not reported)’, ‘yes’  Anxiety ‘no (including not reported)’, ‘yes’  Stress management ‘no (including not reported)’, ‘yes’  Type A behaviour ‘no (including not reported)’, ‘yes’ Psychological components  Relaxation ‘no (including not reported)’, ‘yes’  Cognitive techniques ‘no (including not reported)’, ‘yes’  Emotional support and/or client-led discussion ‘no (including not reported)’, ‘yes’  Adjunct pharmacology ‘no (including not reported)’, ‘yes’ Variable . Levelsa . Targeting of psychological interventions ‘non-selected population (including not reported)’, ‘population with clinically established psychological disorder’ Mode of intervention delivery ‘individual (including not reported)’, ‘group or mix of individual & group’ Family involvement in intervention ‘no (including not reported)’, ‘yes’ Cardiac risk factor education included as part of the intervention ‘no (including not reported)’, ‘yes’ Behaviour change for cardiac risk factors included as part of the intervention ‘no (including not reported)’, ‘yes’ Psychological treatment targets  Depression ‘no (including not reported)’, ‘yes’  Anxiety ‘no (including not reported)’, ‘yes’  Stress management ‘no (including not reported)’, ‘yes’  Type A behaviour ‘no (including not reported)’, ‘yes’ Psychological components  Relaxation ‘no (including not reported)’, ‘yes’  Cognitive techniques ‘no (including not reported)’, ‘yes’  Emotional support and/or client-led discussion ‘no (including not reported)’, ‘yes’  Adjunct pharmacology ‘no (including not reported)’, ‘yes’ a First level coded ‘0’ and second level coded ‘1’ in regression models. Open in new tab All statistical analyses were performed using Review Manager 5.3 Software16 and STATA version 13.0 (StataCorp, College Station, TX, USA).17 Results Selection and inclusion of studies The 2011 review identified 24 studies that met the inclusion criteria. On review, three studies were excluded due to either an ineligible patient population,18 an inappropriate control group19 or a non-randomised trial design20 and therefore 21 of the 24 studies were included in this update. Searches between 2009 and 2016 yielded 6359 titles and abstracts (Supplementary Figure 1). A total of 123 papers were reviewed and 14 studies (2577 participants) met the inclusion criteria.21–34 Thus, a total of 35 studies (81 publications) were included, reporting data from 10,703 participants (Supplementary Table 1 provides a full bibliography). Study, participant and intervention characteristics Studies Most studies were published in Europe (19 studies) or North America (12 studies) (Table 2). While studies randomised between 42 and 2481 participants, most were small, with a median sample size of 123 participants (interquartile range [IQR] 73–204). The median length of follow-up was 12 months (IQR 12–29 months); longer follow-ups (over 30 months) were restricted to clinical events data extracted from routine records rather than psychological outcomes. Table 2. Study, participant and intervention characteristics. Study characteristics (35 studies) . n Studies . Study location  Europe 19  North America 12  Australia 4  China 1  Duration of follow-up, months (range)a 12 (6, 128)  Median sample size (range) 123 (42, 2481)  Median duration of follow-up,  months (range)a 12 (6, 128) Population characteristics  Median of study mean ages, years (range) 59.6 (53–67)  Median proportion of males (range) 77 (0–100) Cardiac indication on referral, %  Myocardial infarction 65.7  Revascularisation procedure 27.4 Psychological disorder present at baseline  All sample (inclusion criterion) 12  Mixed (observed, not required)b 11  None (exclusion criterion) 3  Not reported 11 Intervention characteristics  Settingc  Hospital 9  Clinic 7  Home-based 4  Mixed (inpatient, other support) 2  Not reported 13  Median treatment contact hours (range) 12 (2–96) Mode of delivery  Group 20  Individual (including not reported) 10  Mixed (group/individual) 5 Family involvement with treatment  Yes 11  Not reported 24 Psychological treatment aims/components  Multiple aims/components 23  Single aim/component 12 Treatment aims  Stress 22  Depression 17  Anxiety 16  Type A behaviour (including anger/hostility) 12  Improving disease adjustment 11 Treatment components  Relaxation techniques 20  Self-awareness and self-monitoring 20  Cognitive challenge or restructuring 19  Emotional support or client-led discussion 15 Treatment co-interventions  Behavioural change for cardiac risk factors 19  Awareness of cardiac risk factors 16  Psycho-pharmacological prescribing 3 Study characteristics (35 studies) . n Studies . Study location  Europe 19  North America 12  Australia 4  China 1  Duration of follow-up, months (range)a 12 (6, 128)  Median sample size (range) 123 (42, 2481)  Median duration of follow-up,  months (range)a 12 (6, 128) Population characteristics  Median of study mean ages, years (range) 59.6 (53–67)  Median proportion of males (range) 77 (0–100) Cardiac indication on referral, %  Myocardial infarction 65.7  Revascularisation procedure 27.4 Psychological disorder present at baseline  All sample (inclusion criterion) 12  Mixed (observed, not required)b 11  None (exclusion criterion) 3  Not reported 11 Intervention characteristics  Settingc  Hospital 9  Clinic 7  Home-based 4  Mixed (inpatient, other support) 2  Not reported 13  Median treatment contact hours (range) 12 (2–96) Mode of delivery  Group 20  Individual (including not reported) 10  Mixed (group/individual) 5 Family involvement with treatment  Yes 11  Not reported 24 Psychological treatment aims/components  Multiple aims/components 23  Single aim/component 12 Treatment aims  Stress 22  Depression 17  Anxiety 16  Type A behaviour (including anger/hostility) 12  Improving disease adjustment 11 Treatment components  Relaxation techniques 20  Self-awareness and self-monitoring 20  Cognitive challenge or restructuring 19  Emotional support or client-led discussion 15 Treatment co-interventions  Behavioural change for cardiac risk factors 19  Awareness of cardiac risk factors 16  Psycho-pharmacological prescribing 3 a The length of follow-up of clinical events; psychological outcomes were often followed up for shorter periods within the overall assessment schedule. b Includes two studies in which the inclusion criterion was a confirmed psychopathology and/or another indicating condition. c Clinical settings can include cardiac rehabilitation units, hospital outpatient clinics or community centres. Open in new tab Table 2. Study, participant and intervention characteristics. Study characteristics (35 studies) . n Studies . Study location  Europe 19  North America 12  Australia 4  China 1  Duration of follow-up, months (range)a 12 (6, 128)  Median sample size (range) 123 (42, 2481)  Median duration of follow-up,  months (range)a 12 (6, 128) Population characteristics  Median of study mean ages, years (range) 59.6 (53–67)  Median proportion of males (range) 77 (0–100) Cardiac indication on referral, %  Myocardial infarction 65.7  Revascularisation procedure 27.4 Psychological disorder present at baseline  All sample (inclusion criterion) 12  Mixed (observed, not required)b 11  None (exclusion criterion) 3  Not reported 11 Intervention characteristics  Settingc  Hospital 9  Clinic 7  Home-based 4  Mixed (inpatient, other support) 2  Not reported 13  Median treatment contact hours (range) 12 (2–96) Mode of delivery  Group 20  Individual (including not reported) 10  Mixed (group/individual) 5 Family involvement with treatment  Yes 11  Not reported 24 Psychological treatment aims/components  Multiple aims/components 23  Single aim/component 12 Treatment aims  Stress 22  Depression 17  Anxiety 16  Type A behaviour (including anger/hostility) 12  Improving disease adjustment 11 Treatment components  Relaxation techniques 20  Self-awareness and self-monitoring 20  Cognitive challenge or restructuring 19  Emotional support or client-led discussion 15 Treatment co-interventions  Behavioural change for cardiac risk factors 19  Awareness of cardiac risk factors 16  Psycho-pharmacological prescribing 3 Study characteristics (35 studies) . n Studies . Study location  Europe 19  North America 12  Australia 4  China 1  Duration of follow-up, months (range)a 12 (6, 128)  Median sample size (range) 123 (42, 2481)  Median duration of follow-up,  months (range)a 12 (6, 128) Population characteristics  Median of study mean ages, years (range) 59.6 (53–67)  Median proportion of males (range) 77 (0–100) Cardiac indication on referral, %  Myocardial infarction 65.7  Revascularisation procedure 27.4 Psychological disorder present at baseline  All sample (inclusion criterion) 12  Mixed (observed, not required)b 11  None (exclusion criterion) 3  Not reported 11 Intervention characteristics  Settingc  Hospital 9  Clinic 7  Home-based 4  Mixed (inpatient, other support) 2  Not reported 13  Median treatment contact hours (range) 12 (2–96) Mode of delivery  Group 20  Individual (including not reported) 10  Mixed (group/individual) 5 Family involvement with treatment  Yes 11  Not reported 24 Psychological treatment aims/components  Multiple aims/components 23  Single aim/component 12 Treatment aims  Stress 22  Depression 17  Anxiety 16  Type A behaviour (including anger/hostility) 12  Improving disease adjustment 11 Treatment components  Relaxation techniques 20  Self-awareness and self-monitoring 20  Cognitive challenge or restructuring 19  Emotional support or client-led discussion 15 Treatment co-interventions  Behavioural change for cardiac risk factors 19  Awareness of cardiac risk factors 16  Psycho-pharmacological prescribing 3 a The length of follow-up of clinical events; psychological outcomes were often followed up for shorter periods within the overall assessment schedule. b Includes two studies in which the inclusion criterion was a confirmed psychopathology and/or another indicating condition. c Clinical settings can include cardiac rehabilitation units, hospital outpatient clinics or community centres. Open in new tab Participants The median of study mean ages was 59.6 years and the median proportion of males was 77% (Table 2). The most common cardiac indication upon study referral was an MI (65.7%), with around a third having undergone some form of revascularisation procedure (27.4%). Twelve studies required participants to have a clinical psychopathology (most commonly depression) at baseline to satisfy an eligibility criterion. In unselected cardiac populations, nine studies reported rates of depression of between 3.8%35 and 53%36 and three studies reported rates of anxiety of 32%33,37 and 53%38 (some papers reported both anxiety and depression). Only three excluded individuals with psychopathology at baseline and 11 studies either did not measure psychological outcomes at baseline or did not report them. Interventions The number of contact hours in psychological interventions varied considerably, ranging from an average of 2 hours to 96 hours (31 studies; Table 2). Over half were delivered in groups (20 studies) or a mix of group and individual sessions (five studies). Eleven studies reported family involvement in treatment. Although the quality of reporting of interventions was highly variable, based on available descriptions, 23 studies evaluated psychological treatments with multiple treatment aims and components. Common treatment aims included managing stress (22 studies), depression (17 studies), anxiety (16 studies) and Type A behaviour including anger and hostility (12 studies), as well as achieving improved disease adjustment (11 studies). Common treatment components included relaxation techniques (20 studies), self-awareness and self-monitoring (20 studies), emotional support or client-led discussion (15 studies) and cognitive challenge or cognitive restructuring techniques (19 studies). Many interventions included co-interventions aimed at raising awareness of cardiac risk factors (16 studies) and the targeting of behaviours relating to cardiac risk reduction (e.g. smoking and salt intake; 19 studies). Only three studies incorporated the co-prescribing of pharmacological drugs where it was deemed clinically appropriate.21,29,39 Risk of bias and GRADE assessment The overall risk of bias scores varied between items assessed (Supplementary Table 2). The quality of reporting was highly variable, with an unclear risk of bias for over half the studies for domains relating to randomisation procedures and the blinding of outcome assessment. This limited our ability to judge risk of bias and thus downgrading the GRADE quality of evidence across all outcomes (Tables 3 and 4). Table 3. Results from the pooled analysis of mortality and cardiovascular morbidity. Outcome (median follow-up) . Number of participants (studies) . Number of events . . Statistical heterogeneity I2 (p-value) . GRADE quality of evidence . Intervention . Comparator . RR (95% CI) . Total mortality (13 months) 7776 (23) 319/3899 352/3877 0.90 (0.77, 1.05) 2% (0.43) Moderatea Cardiovascular mortality (57 months) 4792 (11) 140/2561 161/2231 0.79 (0.63, 0.98) 0% (0.76) Lowa,b Revascularisation (CABG/PCI) (12 months) 6822 (13) 395/3429 412/3393 0.94 (0.81, 1.11) 8% (0.36) Moderatea Non-fatal MI (30 months) 7845 (13) 340/4114 355/3731 0.82 (0.64, 1.05) 41% (0.07) Lowa,c Outcome (median follow-up) . Number of participants (studies) . Number of events . . Statistical heterogeneity I2 (p-value) . GRADE quality of evidence . Intervention . Comparator . RR (95% CI) . Total mortality (13 months) 7776 (23) 319/3899 352/3877 0.90 (0.77, 1.05) 2% (0.43) Moderatea Cardiovascular mortality (57 months) 4792 (11) 140/2561 161/2231 0.79 (0.63, 0.98) 0% (0.76) Lowa,b Revascularisation (CABG/PCI) (12 months) 6822 (13) 395/3429 412/3393 0.94 (0.81, 1.11) 8% (0.36) Moderatea Non-fatal MI (30 months) 7845 (13) 340/4114 355/3731 0.82 (0.64, 1.05) 41% (0.07) Lowa,c a Random sequence generation, allocation concealment or blinding of outcome assessors poorly described in ≥50% of included studies. b Egger tests suggest evidence of asymmetry. c 95% CIs include both no effect and appreciable benefit or harm (i.e. 95% CI < 0.75 or >1.25). GRADE: moderate = further research is very likely to have an important effect on confidence of the estimated effect and may change the estimate; low = further research is very likely to have an important effect on confidence in the estimated effect and is likely to change the estimate. CABG: coronary artery bypass grafting; CI: confidence interval; GRADE: Grading of Recommendations Assessment, Development and Evaluation; MI: myocardial infarction; PCI: percutaneous coronary intervention; RR: relative risk. Open in new tab Table 3. Results from the pooled analysis of mortality and cardiovascular morbidity. Outcome (median follow-up) . Number of participants (studies) . Number of events . . Statistical heterogeneity I2 (p-value) . GRADE quality of evidence . Intervention . Comparator . RR (95% CI) . Total mortality (13 months) 7776 (23) 319/3899 352/3877 0.90 (0.77, 1.05) 2% (0.43) Moderatea Cardiovascular mortality (57 months) 4792 (11) 140/2561 161/2231 0.79 (0.63, 0.98) 0% (0.76) Lowa,b Revascularisation (CABG/PCI) (12 months) 6822 (13) 395/3429 412/3393 0.94 (0.81, 1.11) 8% (0.36) Moderatea Non-fatal MI (30 months) 7845 (13) 340/4114 355/3731 0.82 (0.64, 1.05) 41% (0.07) Lowa,c Outcome (median follow-up) . Number of participants (studies) . Number of events . . Statistical heterogeneity I2 (p-value) . GRADE quality of evidence . Intervention . Comparator . RR (95% CI) . Total mortality (13 months) 7776 (23) 319/3899 352/3877 0.90 (0.77, 1.05) 2% (0.43) Moderatea Cardiovascular mortality (57 months) 4792 (11) 140/2561 161/2231 0.79 (0.63, 0.98) 0% (0.76) Lowa,b Revascularisation (CABG/PCI) (12 months) 6822 (13) 395/3429 412/3393 0.94 (0.81, 1.11) 8% (0.36) Moderatea Non-fatal MI (30 months) 7845 (13) 340/4114 355/3731 0.82 (0.64, 1.05) 41% (0.07) Lowa,c a Random sequence generation, allocation concealment or blinding of outcome assessors poorly described in ≥50% of included studies. b Egger tests suggest evidence of asymmetry. c 95% CIs include both no effect and appreciable benefit or harm (i.e. 95% CI < 0.75 or >1.25). GRADE: moderate = further research is very likely to have an important effect on confidence of the estimated effect and may change the estimate; low = further research is very likely to have an important effect on confidence in the estimated effect and is likely to change the estimate. CABG: coronary artery bypass grafting; CI: confidence interval; GRADE: Grading of Recommendations Assessment, Development and Evaluation; MI: myocardial infarction; PCI: percutaneous coronary intervention; RR: relative risk. Open in new tab Table 4. Results from the pooled analysis psychological outcomes. Outcome (median follow-up) . Number of participants (studies) . SMD (95% CI) (intervention – comparator) . Statistical heterogeneity I2 (p-value) . GRADE quality of evidence . Depression (12 months) 5829 (19) –0.27 (–0.39, –0.15) 69% (<0.001) ++– Lowa,b Anxiety (12 months) 3165 (12) –0.24 (–0.38, –0.09) 47% (0.03) ++– Lowa,c Stress (12 months) 1255 (8) –0.56 (–0.88, –0.24) 86% (<0.001) +— Very lowa,b,d Outcome (median follow-up) . Number of participants (studies) . SMD (95% CI) (intervention – comparator) . Statistical heterogeneity I2 (p-value) . GRADE quality of evidence . Depression (12 months) 5829 (19) –0.27 (–0.39, –0.15) 69% (<0.001) ++– Lowa,b Anxiety (12 months) 3165 (12) –0.24 (–0.38, –0.09) 47% (0.03) ++– Lowa,c Stress (12 months) 1255 (8) –0.56 (–0.88, –0.24) 86% (<0.001) +— Very lowa,b,d a Random sequence generation, allocation concealment or blinding of outcome assessors were poorly described in ≥50% of included studies. b Moderate heterogeneity (I2 > 50%). c Egger tests suggest evidence of asymmetry. d 95% CIs around the SMD did not include the value of a + 5 at either the lower or upper limits (indicative of clinical significance). GRADE: moderate = further research is very likely to have an important effect on confidence of the estimated effect and may change the estimate; low = further research is very likely to have an important effect on confidence in the estimated effect and is likely to change the estimate; very low quality = the estimate is very uncertain. CI: confidence interval; GRADE: Grading of Recommendations Assessment, Development and Evaluation; SMD: standardised mean difference. Open in new tab Table 4. Results from the pooled analysis psychological outcomes. Outcome (median follow-up) . Number of participants (studies) . SMD (95% CI) (intervention – comparator) . Statistical heterogeneity I2 (p-value) . GRADE quality of evidence . Depression (12 months) 5829 (19) –0.27 (–0.39, –0.15) 69% (<0.001) ++– Lowa,b Anxiety (12 months) 3165 (12) –0.24 (–0.38, –0.09) 47% (0.03) ++– Lowa,c Stress (12 months) 1255 (8) –0.56 (–0.88, –0.24) 86% (<0.001) +— Very lowa,b,d Outcome (median follow-up) . Number of participants (studies) . SMD (95% CI) (intervention – comparator) . Statistical heterogeneity I2 (p-value) . GRADE quality of evidence . Depression (12 months) 5829 (19) –0.27 (–0.39, –0.15) 69% (<0.001) ++– Lowa,b Anxiety (12 months) 3165 (12) –0.24 (–0.38, –0.09) 47% (0.03) ++– Lowa,c Stress (12 months) 1255 (8) –0.56 (–0.88, –0.24) 86% (<0.001) +— Very lowa,b,d a Random sequence generation, allocation concealment or blinding of outcome assessors were poorly described in ≥50% of included studies. b Moderate heterogeneity (I2 > 50%). c Egger tests suggest evidence of asymmetry. d 95% CIs around the SMD did not include the value of a + 5 at either the lower or upper limits (indicative of clinical significance). GRADE: moderate = further research is very likely to have an important effect on confidence of the estimated effect and may change the estimate; low = further research is very likely to have an important effect on confidence in the estimated effect and is likely to change the estimate; very low quality = the estimate is very uncertain. CI: confidence interval; GRADE: Grading of Recommendations Assessment, Development and Evaluation; SMD: standardised mean difference. Open in new tab Some outcomes were also downgraded due to a lack of precision around the estimated effect (non-fatal MI and stress), significant heterogeneity observed (anxiety and stress) and/or the risk of publication bias (cardiac mortality and anxiety). Thus, the GRADE ratings were moderate (total mortality and revascularisation), low (cardiac mortality, non-fatal MI, anxiety and depression) or very low (stress) for all outcomes. Outcome results For mortality and cardiovascular morbidity data, the attrition at follow-up was low with, for example, 1.7% of total mortality data missing from the pooled analysis of 23 studies. In contrast, the overall level of attrition of studies contributing to the pooled analyses was 17.7% for depression, 9.1% for anxiety and 9.4% for stress. Mortality Pooled analysis of 23 studies (Table 3, Supplementary Figure 2) found no evidence that psychological therapies reduced the risk of total mortality (7776 participants; RR 0.90, 95% CI 0.77 to 1.05, I2 = 2%). However, there was evidence that psychological interventions reduced the risk of cardiac mortality (Table 3, Figure 1) when pooling data from 11 studies (4792 participants, RR 0.79, 95% CI 0.63 to 0.98, I2 = 0%), although there is some uncertainty in this finding as the quality of evidence is low. Figure 1. Open in new tabDownload slide Forest plot of psychological intervention versus usual care: cardiovascular mortality. CI: confidence interval; df: degrees of freedom; M-H: Mantel-Haenszel method. Reproduced from Richards et al.10 Cardiovascular morbidity There was no evidence of a risk reduction for revascularisation procedures (Table 3, Supplementary Figure 3) (13 studies, 6822 participants; RR 0.94, 95% CI 0.81 to 1.11, I2 = 8%) or for an occurrence of a subsequent non-fatal MI (Table 3, Supplementary Figure 4) (13 studies, 7845 participants; 0.82, 95% CI 0.64 to 1.05, I2 = 41%). Psychological outcomes A meta-analysis of 19 studies (5825 participants) found evidence that psychological interventions reduced depression symptoms compared with the comparator group (SMD –0.27, 95% CI –0.39 to –0.15, I2 = 69%; Table 4, Figure 2). Reductions in anxiety levels (12 trials, 3161 participants; SMD –0.24, 95% CI –0.38 to –0.09, I2 = 47%; Table 4, Figure 3) and stress levels (eight trials, 1251 participants; SMD –0.56, 95% CI –0.88 to –0.24, I2 = 86%; Table 4, Figure 4) in favour of the intervention group were also observed. However, there remains considerable uncertainty regarding treatment effects for all comparisons as the quality of evidence was either low or very low (Table 4). Figure 2. Open in new tabDownload slide Forest plot of psychological intervention versus usual care: depression. CI: confidence interval; df: degrees of freedom; IV: inverse variance method. Reproduced from Richards et al.10 Figure 3. Open in new tabDownload slide Forest plot of psychological intervention versus usual care: anxiety. CI: confidence interval; df: degrees of freedom; IV: inverse variance method. Reproduced from Richards et al.10 Figure 4. Open in new tabDownload slide Forest plot of psychological intervention versus usual care: stress. CI: confidence interval; df: degrees of freedom; IV: inverse variance method. Reproduced from Richards et al.10 Statistical heterogeneity and small study bias Inspection of I2 tests found significant levels of statistical heterogeneity in the meta-analyses of all psychological outcomes, but not mortality or morbidity data. Visual inspection of the funnel plots (data reported elsewhere10) shows some evidence of asymmetry for cardiac mortality, depression, anxiety and stress, but not total mortality or other measures of cardiovascular morbidity. The Egger tests for funnel plot asymmetry were non-significant for the majority of primary outcomes, with the exceptions of cardiovascular mortality (p = 0.04) and anxiety (p = 0.012). This asymmetry appeared to be due to an absence of small- to medium-sized studies with negative results regarding psychological interventions. Health-related quality of life HRQL was reported in ten studies (Supplementary Table 3). A narrative review found statistically significant improvements in at least one dimension of HRQL in favour of psychological interventions in four studies,22,28,29,40 while six studies26,33,35,41–43 reported no between-group differences. Of studies reporting significant treatment effects, two observed improvements restricted to mental health and/or life satisfaction components of HRQL22,40 and a third study found improvements restricted to the physical health component,29 while the fourth study reported improvements in both physical and mental health components.28 Cost-effectiveness Only two studies reported any form of economic evaluation alongside trial data. Van-Dixhoorn and Duivenvoorden44 limited the economic evaluation to an examination of hospital costs arising from cardiac-related hospital readmissions across a 5-year follow-up. The authors reported that the extra costs of individual relaxation training sessions (the intervention) were outweighed by the benefits (30% reduction in the number of days in hospital and 46% reduction in costs due to reduced readmissions for cardiac surgery). Davidson et al.21 (see Ladapo et al.45) examined HRQL, health care utilisation and costs of the intervention compared to usual physician care. The mean total health care costs (psychotropic medicines, ambulatory care and hospitalisations) were $1857 for the intervention group and $2797 for the usual care group (adjusted difference –$1229 per patient, 95% CI –$2652 to $195, p = 0.09), with a 98% probability that this approach would be considered cost effective if a willingness-to-pay threshold of $30,000 per quality-adjusted life-year gained was applied. Meta-regression findings We found no significant predictors of intervention effects for total or cardiovascular mortality (Supplementary Table 4) for any of the population or intervention characteristics explored in univariate meta-regression models. Meta-regression of psychological outcomes yielded only two statistically significant predictor variables. Psychological interventions combined with pharmacology for an underlying psychological disorder (p = 0.003) were more effective at alleviating depression than interventions that were not (Supplementary Table 4). Interventions recruiting participants with an underlying psychological disorder were more effective at alleviating anxiety than those delivered to unselected populations (p = 0.03). Discussion Main findings We updated a systematic review of the direct effects of psychological interventions for people with CHD. We found a reduction in cardiovascular mortality (7.3% to 5.5%; number needed to treat: 56) with psychological interventions compared with usual care controls. No between-group differences were observed for the rates of total mortality, non-fatal MI or revascularisation procedures. Psychological interventions were found to achieve small to moderate improvements in depressive symptoms, anxiety and stress compared with controls, although there remains some uncertainty in these estimates. A narrative synthesis found some evidence of a positive effect on HRQL, although direct comparisons are problematic due to methodological differences between studies, such as the use of different HRQL measures. Only two studies conducted economic evaluations, with both concluding that psychological therapies were likely to be cost effective, although this evidence requires replication in future research. We undertook an exploratory analysis seeking to identify potential effect modifiers from a range of population and intervention characteristics. In contrast to the previous update,9 we elected not analyse some the patient characteristics of study populations (e.g. sex or age) previously explored using meta-regression. Recent methodological guidance for systematic reviews of cardiac prevention studies46 cautions against the analysis of patient characteristics in meta-regression when aggregated at the study level. Statistically, study-level analysis is underpowered compared with individual patient data meta-analysis. More importantly, however, this form of analysis is prone to ecological fallacy (or ‘aggregation bias’). Meta-regression failed to identify any predictor variables for total and cardiovascular mortality, although this was not unexpected given the lack of statistical heterogeneity in the pooled analysis. Meta-regression for the outcomes of depression and anxiety, where considerably greater statistical heterogeneity was observed in pooled analysis, found only two predictor variables. For depression, the adjunct use of pharmacological therapy for the underlying psychological condition (where deemed clinically appropriate) may increase intervention effectiveness compared with interventions that did not. For anxiety, psychological interventions that recruited participants with CHD and an underlying psychological disorder appeared more effective than those delivered to unselected CHD populations. Findings in context Our study has further clarified findings from the 2011 update,9 with the precision of the effect estimates improving across all outcomes through the inclusion of new data from 14 studies (2577 participants). We also present pooled data on stress levels for the first time. However, the meta-regression failed to replicate the effect modifiers (e.g. interventions targeting Type A behaviours or involving family members) previously identified for the outcome of depression. This is likely to be attributable to the inclusion of a number of new studies combined with the exclusion of data from two studies that had previously contributed data to these analyses.19,47 Although other systematic reviews have sought to explore the effectiveness of psychological interventions for people with CHD,48,49 direct comparisons are problematic due to important differences in study selection. For example, Welton et al.48 included studies testing both the direct and indirect effects of psychological interventions for people with CHD, whilst Dickens et al.49 included studies with a follow-up period of less than 6 months. In contrast to our findings, Welton et al.48 found no evidence that psychological interventions reduced cardiovascular mortality, although, consistent with our findings, no effect on total mortality was observed. There is also consistent evidence emerging across a body of empirical evidence showing that psychological interventions have small but consistent effects at alleviating symptoms of depression48,49 and anxiety48 for people with CHD. Notwithstanding the uncertainty regarding the optimal methods of providing psychological care, this review lends further support to the international guidelines4,6–8 suggesting that addressing psychological health should be a core component of conventional cardiac prevention services. Study limitations The level of reporting of key risk of bias domains relating to randomisation procedures and the blinding of outcome assessment was poor, limiting our ability to judge risk of bias. Some outcomes were also downgraded due to a lack of precision around the estimated effect, significant heterogeneity observed and/or the risk of publication bias. Thus, the GRADE quality of evidence was moderate, low or very low across outcomes. From the information reported, the majority of participants were men recruited post-MI and so our findings may be less generalisable to more diverse populations of women or to individuals with other cardiac conditions using secondary prevention services. Another feature of the studies synthesised was the clinical heterogeneity, as studies often tested complex psychological interventions with multiple treatment targets and components; only a minority tested the effectiveness of single-component therapies (e.g. Van-Dixhoorn and Duivenvoorden44 and Blumenthal et al.34 tested a stress management intervention). The poor reporting of intervention components (e.g. the training received or any ongoing supervision provided) and participant characteristics (e.g. a third of studies did not report the presence of psychopathology at baseline) limited a detailed examination of the active ingredients of psychological techniques through meta-regression. While meta-analysis found evidence of small effects on a number of outcomes, there remains considerable uncertainty regarding which type of psychological techniques are most effective and for whom. The effectiveness of emerging and potentially more beneficial psychological interventions has yet to be addressed: mindfulness, for example, may be more effective than traditional stress management approaches for individuals with high levels of health anxiety.50 In addition, given the likely low effect sizes (in terms of both psychological and cardiac benefit) of any psychological interventions targeted at a population with no obvious psychopathology, testing the effectiveness of interventions for people with established psychopathology is an important issue to address in future studies. A number of ongoing trials appear to be directly assessing some of these uncertainties.51–53 Our review also excluded psychological interventions designed specifically to improve adherence to cardiac risk factor modification (e.g. medicines and lifestyle change); this was essential in order to reduce the clinical heterogeneity of the compared interventions, but as a consequence, our findings do not inform the wider evidence base on the contribution of psychological techniques to optimising risk factor management. While we were able to pool data for a number of important clinical and psychological outcomes, the breadth of outcome measures reported was often limited within studies. For example, while around two-thirds of studies (23/35) reported total mortality, less than a third of studies reported stress levels (8/35) or cardiovascular mortality (11/35) in a way that could be pooled. In addition, the reporting of the psychological status of the study populations at baseline was often omitted and only a minority of studies reported other important outcomes, such as HRQL or data that could be used to support health economic evaluation. Conclusions This updated Cochrane review found that psychological treatments had important health benefits among people with CHD, reducing the rate of cardiac mortality and alleviating the psychological symptoms of depression, anxiety and stress. However, according to the GRADE methodology, there remains uncertainty regarding these benefits and large-scale trials are still warranted. Future trials must provide a clearer reporting of their methods and interventions (perhaps following similar taxonomies of intervention components to those encouraged in health behaviour interventions54), assess a broader range of outcomes and undertake health economic evaluation. There also remains uncertainty regarding who benefits most from treatment and which types of psychological intervention yield the greatest benefit. Future trials that test the efficacy of specific psychological techniques are still needed, although this may prove challenging in real-world settings where patients may present with complex psychological needs that alter across the course of their recovery. Pragmatic trials of multifactorial interventions, delivered in a blended fashion, are also justified, but should be accompanied by pre-planned process evaluations (e.g. using subgroup analysis) in order to better understand the active ingredients of such complex interventions.55 Future trials should also explore the optimal targeting of interventions for people with CHD with or without psychopathologies. Author contribution RST and SHR contributed to the conception and design of this review, building on the work undertaken by the authors of the two previous versions (see below). SHR, LA and CEJ undertook study selection, data acquisition, data extraction and risk of bias assessment. SHR, LA and RST undertook data analysis. KR was the lead author on the first version (2004) of this review and a co-author on the second version (2011). BW was the lead author on the second version (2011) of this review and in this third update advised with study selection and analyses and provided advice on classifying study interventions. PB and RW were co-authors on both the first and second versions of this review. RST, PD, ZL and DRT were co-authors on the second version of this review. All authors edited the manuscript, gave final approval and agree to be accountable for all aspects of the work ensuring integrity and accuracy. Acknowledgements The authors thank Cornelia Junghans, Jerong Ji and Mensrain Mujeeb for their translation services and Linda Long for her assistance with data checking. We also thank all of the authors who provided additional information about their trials and the co-authors of the two previous versions of this review. Finally, we thank the Cochrane Heart Group for their support of the co-publication of this article with the full version of the review, which is published on the Cochrane Database of Systematic Reviews. This paper is a synthesis of a previously published systematic review by the Cochrane Collaboration: Richards SH, Anderson L, Jenkinson CE, Whalley B, Rees K, Davies P, Bennett P, Liu Z, West R, Thompson DR and Taylor RS. Psychological interventions for coronary heart disease. Cochrane Database Systematic Reviews 2017; 4: CD002902. Declaration of conflicting interests The author(s) declared the following potential conflicts of interest with respect to the research, authorship and/or publication of this article: SHR is currently a co-investigator on the CADENCE study (funded by the UK NHIR HTA 12/189/06). This study is a feasibility and pilot study aimed at developing enhanced psychological care for people with new-onset depression using cardiac rehabilitation services (ISRCTN34701576). KR, PB and RW were the authors of the first version (2004) of this review. BW, KR, PD, PB, ZL, RW, DRT and RST were the authors of the second version (2011) of this review. KR, DRT, LA and RST are the authors of a number of other Cochrane cardiac rehabilitation reviews. RST is currently the co-chief investigator on the programme of research with the overarching aims of developing and evaluating a home-based cardiac rehabilitation intervention for people with heart failure and their carers (UK NIHR PGfAR RP-PG-0611-12004). RST is also currently a co-investigator on the CADENCE study (funded by the UK NHIR HTA 12/189/06). The other authors declare no other conflicts of interest. Funding The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: this study was supported by a small grant from the UK South West General Practice Trust (registered charity 292013). All authors have been supported by funding from their host universities. In addition, RST received support from the UK National Institute for Health Research (NIHR) Collaboration for Leadership in Applied Health Research and Care South West Peninsula at the Royal Devon and Exeter NHS Foundation Trust, UK. BW’s input was supported during the second update by a postdoctoral fellowship (PTA-026-27-2113) from the UK Economic and Social Science Research Council. The UK NIHR Health Technology Assessment Programme CADENCE Study (12/189/06) also supported SHR and RST. The South West General Practice Trust (registered charity 292013), UK, provided a small project award (chief investigator SHR) to support CEJ and LA’s contributions. Cochrane Infrastructure funding to the Heart Group UK supported production of the Cochrane review. The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the NHS, the UK NIHR or the UK Department of Health. References 1 World Health Organisation. The top 10 causes of death, 2014. http://www.who.int/mediacentre/factsheets/fs310/en . 2 Tully PJ , Cosh SM, Baumeister H. The anxious heart in whose mind? A systematic review and meta-regression of factors associated with anxiety disorder diagnosis, treatment and morbidity risk in coronary heart disease . J Psychosom Res 2014 ; 77 : 439 – 448 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Dickens C . Depression in people with coronary heart disease: Prognostic significance and mechanisms . Curr Cardiol Rep 2015 ; 17 : 83 – 83 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Lichtman JH , Froelicher ES, Blumenthal JAet al. Depression as a risk factor for poor prognosis among patients with acute coronary syndrome: Systematic review and recommendations: A scientific statement from the American Heart Association . Circulation 2014 ; 129 : 1350 – 1369 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Gale CR , Batty D, Osborn DPJet al. Mental disorders across the adult life course and future coronary heart disease . Circulation 2014 ; 129 : 186 – 193 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Woodruffe S , Neubeck L, Clark RAet al. Australian Cardiovascular Health and Rehabilitation Association (ACRA) core components of cardiovascular disease secondary prevention and cardiac rehabilitation 2014 . Heart Lung Circ 2015 ; 24 : 430 – 441 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Piepoli MF , Corrà U, Adamopoulos Set al. Secondary prevention in the clinical management of patients with cardiovascular diseases. Core components, standards and outcome measures for referral and delivery. A policy statement from the Cardiac Rehabilitation Section of the European Association for Cardiovascular Prevention & Rehabilitation . Eur J Prev Cardiol 2014 ; 21 : 664 – 681 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Authors/Task Force Members; , Piepoli MF, Hoes AW, Agewall Set al. 2016 European Guidelines on cardiovascular disease prevention in clinical practice: The Sixth Joint Task Force of the European Society of Cardiology and Other Societies on Cardiovascular Disease Prevention in Clinical Practice (constituted by representatives of 10 societies and by invited experts): Developed with the special contribution of the European Association for Cardiovascular Prevention & Rehabilitation (EACPR) . Eur J Prev Cardiol 2016 ; 23 : NP1 – NP96 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Whalley B , Rees K, Davies Pet al. Psychological interventions for coronary heart disease . Cochrane Database Syst Rev 2011 ; 8 : CD002902 – CD002902 . Google Scholar OpenURL Placeholder Text WorldCat 10 Richards SH , Anderson L, Jenkinson CEet al. Psychological interventions for coronary heart disease . Cochrane Database Syst Rev 2017 ; 4 : CD002902 – CD002902 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 11 Higgins J , Green S. Cochrane Handbook for Systematic Reviews of Interventions 2011 . Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. Available at www.handbook.cochrane.org . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 12 Moher D , Liberati A, Tetzlaff Jet al. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement . BMJ 2009 ; 339 : b2535 – b2535 . Google Scholar Crossref Search ADS PubMed WorldCat 13 GRADEpro GTD. GRADEpro Guideline Development Tool [software], 2015. www.gradepro.org . 14 Egger M , Davey Smith G, Schneider Met al. Bias in meta-analysis detected by a simple graphical test . BMJ 1997 ; 315 : 629 – 634 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Deeks JJ , Higgins JPT, Altman DG Chapter 9: Analysing data and undertaking meta-analyses . In: Higgins J, Green S (eds). Cochrane Handbook for Systematic Reviews of Interventions 2011 . Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. Available at www.handbook.cochrane.org . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 16 The Cochrane Community. Review Manager (RevMan). 5.3 Edition. Copenhagen, Denmark: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014. 17 StataCorp LP . STATA Software (Version 14.1) , College Station, TX, USA : StataCorp , 2016 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 18 Cowan MJ , Pike KC, Budzynski HK. Psychosocial nursing therapy following sudden cardiac arrest: Impact on two-year survival . Nurs Res 2001 ; 50 : 68 – 76 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Hofman-Bang C , Lisspers J, Nordlander Ret al. Two-year results of a controlled study of residential rehabilitation for patients treated with percutaneous transluminal coronary angioplasty. A randomized study of a multifactorial programme . Eur Heart J 1999 ; 20 : 1465 – 1474 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Ibrahim MA , Feldman JG, Sultz HAet al. Management after myocardial infarction: A controlled trial of the effect of group psychotherapy . Int J Psychiatry Med 1974 ; 5 : 253 – 268 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Davidson KW , Rieckmann N, Clemow Let al. Enhanced depression care for patients with acute coronary syndrome and persistent depressive symptoms: Coronary psychosocial evaluation studies randomized controlled trial . Arch Intern Med 2010 ; 170 : 600 – 608 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Freedland KE , Skala JA, Carney RMet al. Treatment of depression after coronary artery bypass surgery: A randomized controlled trial . Arch Gen Psychiatry 2009 ; 66 : 387 – 396 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Gulliksson M , Burell G, Wessby Bet al. Randomized controlled trial of cognitive behavioral therapy vs standard treatment to prevent recurrent cardiovascular events in patients with coronary heart disease . Eur Heart J 2011 ; 32 : 390 – 390 . Google Scholar OpenURL Placeholder Text WorldCat 24 Merswolken M , Siebenhuener S, Orth-Gomer Ket al. Treating anxiety in patients with coronary heart disease: A randomized controlled trial . Psychother Psychosom 2011 ; 80 : 365 – 370 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Neves A , Alves AJ, Ribeiro Fet al. The effect of cardiac rehabilitation with relaxation therapy on psychological, hemodynamic, and hospital admission outcome variables . J Cardiopulm Rehabil Prev 2009 ; 29 : 304 – 309 . Google Scholar Crossref Search ADS PubMed WorldCat 26 O’Neil A , Taylor B, Hare DLet al. Long-term efficacy of a tele-health intervention for acute coronary syndrome patients with depression: 12-month results of the MoodCare randomized controlled trial . Eur J Prev Cardiol 2015 ; 22 : 1111 – 1120 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Oranta O , Luutonen S, Salokangas RKet al. The outcomes of interpersonal counselling on depressive symptoms and distress after myocardial infarction . Nord J Psychiatry 2010 ; 64 : 78 – 86 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Rakowska JM . Brief strategic therapy in first myocardial infarction patients with increased levels of stress: A randomized clinical trial . Anxiety Stress Coping 2015 ; 28 : 687 – 705 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Roncella A , Pristipino C, Cianfrocca Cet al. One-year results of the randomized, controlled, short-term psychotherapy in acute myocardial infarction (STEP-IN-AMI) trial . Int J Cardiol 2013 ; 170 : 132 – 139 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Schneider RH , Grim CE, Rainforth MVet al. Stress reduction in the secondary prevention of cardiovascular disease: Randomized, controlled trial of transcendental meditation and health education in blacks . Circ Cardiovasc Qual Outcomes 2012 ; 5 : 750 – 758 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Turner A , Murphy BM, Higgins ROet al. An integrated secondary prevention group programme reduces depression in cardiac patients . Eur J Prev Cardiol 2014 ; 21 : 153 – 162 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Turner A , Hambridge J, Baker Aet al. Randomised controlled trial of group cognitive behaviour therapy versus brief intervention for depression in cardiac patients . Aust N Z J Psychiatry 2013 ; 47 : 235 – 243 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Lie I , Arnesen H, Sandvik Let al. Effects of a home-based intervention program on anxiety and depression 6 months after coronary artery bypass grafting: A randomized controlled trial . J Psychosom Res 2007 ; 62 : 411 – 418 . Google Scholar Crossref Search ADS PubMed WorldCat 34 Blumenthal JA , Sherwood A, Smith PJet al. Enhancing cardiac rehabilitation with stress management training. A randomized, clinical efficacy trial . Circulation 2016 ; 133 : 1341 – 1350 . Google Scholar Crossref Search ADS PubMed WorldCat 35 Michalsen A , Grossman P, Lehmann Net al. Psychological and quality-of-life outcomes from a comprehensive stress reduction and lifestyle program in patients with coronary artery disease: Results of a randomized trial . Psychother Psychosom 2005 ; 74 : 344 – 352 . Google Scholar Crossref Search ADS PubMed WorldCat 36 Koertge J , Janszky I, Sundin Oet al. Effects of a stress management program on vital exhaustion and depression in women with coronary heart disease: A randomized controlled intervention study . J Intern Med 2008 ; 263 : 281 – 293 . Google Scholar Crossref Search ADS PubMed WorldCat 37 Jones DA , West RR. Psychological rehabilitation after myocardial infarction: Multicentre randomised controlled trial . BMJ 1996 ; 313 : 1517 – 1521 . Google Scholar Crossref Search ADS PubMed WorldCat 38 McLaughlin T , Aupont O, Bambauer Ket al. Improving psychologic adjustment to chronic illness in cardiac patients: The role of depression and anxiety . J Gen Intern Med 2005 ; 20 : 1084 – 1090 . Google Scholar Crossref Search ADS PubMed WorldCat 39 Black JL , Allison TG, Williams DEet al. Effect of intervention for psychological distress on rehospitalization rates in cardiac rehabilitation patients . Psychosomatics 1998 ; 39 : 134 – 143 . Google Scholar Crossref Search ADS PubMed WorldCat 40 The ENRICHD Investigators . Enhancing recovery in coronary heart disease patients (ENRICHD): Study design and methods . Am Heart J 2000 ; 139 : 1 – 9 . Crossref Search ADS PubMed WorldCat 41 Appels A , Bär F, van der Pol Get al. Effects of treating exhaustion in angioplasty patients on new coronary events: Results of the randomized exhaustion intervention trial (EXIT) . Psychosom Med 2005 ; 67 : 217 – 223 . Google Scholar Crossref Search ADS PubMed WorldCat 42 Claesson M , Birgander L, Lindahl Bet al. Women’s hearts: Stress management for women with ischemic heart disease: Explanatory analyses of a randomized controlled trial . J Cardiopulm Rehabil 2005 ; 25 : 93 – 102 . Google Scholar Crossref Search ADS PubMed WorldCat 43 Mayou R , Thompson D, Clements Aet al. Guideline-based early rehabilitation after myocardial infarction: A pragmatic randomised controlled trial . J Psychosom Res 2002 ; 52 : 89 – 95 . Google Scholar Crossref Search ADS PubMed WorldCat 44 Van-Dixhoorn JJ , Duivenvoorden HJ. Effect of relaxation therapy on cardiac events after myocardial infarction: A 5-year follow-up study . J Cardiopulm Rehabil 1999 ; 19 : 178 – 185 . Google Scholar Crossref Search ADS PubMed WorldCat 45 Ladapo JA , Shaffer JA, Fang Yet al. Cost-effectiveness of enhanced depression care after acute coronary syndrome: Results from the Coronary Psychosocial Evaluation Studies randomized controlled trial . Arch Intern Med 2012 ; 172 : 1682 – 1684 . Google Scholar Crossref Search ADS PubMed WorldCat 46 Rao G , Lopez-Jimenez F, Boyd Jet al. Methodological standards for meta-analyses and qualitative systematic reviews of cardiac prevention and treatment studies: A scientific statement from the American Heart Association . Circulation 2017 ; 136 : e172 – e194 . Google Scholar Crossref Search ADS PubMed WorldCat 47 Peng J , Jiang LJ. Psychotherapy on negative emotions for the incidence of ischemia-related events in patients with coronary heart disease . Chinese J Clin Rehabil 2005 ; 9 : 38 – 39 . Google Scholar OpenURL Placeholder Text WorldCat 48 Welton NJ , Caldwell DM, Adamopoulos Eet al. Mixed treatment comparison meta-analysis of complex interventions: Psychological interventions in coronary heart disease . Am J Epidemiol 2009 ; 169 : 1158 – 1165 . Google Scholar Crossref Search ADS PubMed WorldCat 49 Dickens C , Cherrington A, Adeyemi Iet al. Characteristics of psychological interventions that improve depression in people with coronary heart disease: A systematic review and meta-regression . Psychosom Med 2013 ; 72 : 211 – 221 . Google Scholar Crossref Search ADS WorldCat 50 McManus F , Surawy C, Muse Ket al. A randomized clinical trial of mindfulness-based cognitive therapy versus unrestricted services for health anxiety (hypochondriasis) . J Consult Clin Psychol 2012 ; 80 : 817 – 828 . Google Scholar Crossref Search ADS PubMed WorldCat 51 Norlund F , Olsson EM, Burell Get al. Treatment of depression and anxiety with internet-based cognitive behavior therapy in patients with a recent myocardial infarction (U-CARE Heart): Study protocol for a randomized controlled trial . Trials 2015 ; 16 : 154 – 154 . Google Scholar Crossref Search ADS PubMed WorldCat 52 Richards SH , Dickens C, Anderson Ret al. Assessing the effectiveness of enhanced psychological care for patients with depressive symptoms attending cardiac rehabilitation compared with treatment as usual (CADENCE): Study protocol for a pilot cluster randomised controlled trial . Trials 2016 ; 17 : 1 – 1 . Google Scholar Crossref Search ADS PubMed WorldCat 53 Spatola CA , Manzoni GM, Castelnuovo Get al. The ACTonHEART study: Rationale and design of a randomized controlled clinical trial comparing a brief intervention based on acceptance and commitment therapy to usual secondary prevention care of coronary heart disease . Health Qual Life Outcomes 2014 ; 12 : 22 – 22 . Google Scholar Crossref Search ADS PubMed WorldCat 54 Michie S , Richardson M, Johnston Met al. The behavior change technique taxonomy (v1) of 93 hierarchically clustered techniques: Building an international consensus for the reporting of behavior change interventions . Ann Behav Med 2013 ; 46 : 81 – 95 . Google Scholar Crossref Search ADS PubMed WorldCat 55 Moore GF , Audrey S, Barker Met al. Process evaluation of complex interventions: Medical Research Council guidance . BMJ 2015 ; 350 : h1258 – h1258 . Google Scholar Crossref Search ADS PubMed WorldCat © The European Society of Cardiology 2018 This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) © The European Society of Cardiology 2018