European Journal of Preventive Cardiology

Abstract Background Previous studies demonstrated that lower outdoor temperatures increase the levels of established cardiovascular disease risk factors, such as blood pressure and lipids. Whether or not low temperatures increase novel cardiovascular disease risk factors levels is not well studied. The aim was to investigate associations of outdoor temperature with a comprehensive range of established and novel cardiovascular disease risk factors in two large Northern European studies of older adults, in whom cardiovascular disease risk is increased. Design and methods Data came from the British Regional Heart Study (4252 men aged 60–79 years) and the Prospective Study of Pravastatin in the Elderly at Risk (5804 men and women aged 70–82 years). Associations between outdoor temperature and cardiovascular disease risk factors were quantified in each study and then pooled using a random effects model. Results With a 5℃ lower mean temperature, total cholesterol was 0.04 mmol/l (95% confidence interval (CI) 0.02–0.07) higher, low density lipoprotein cholesterol was 0.02 mmol/l (95% CI 0.01–0.05) higher and SBP was 1.12 mm Hg (95% CI 0.60–1.64) higher. Among novel cardiovascular disease risk factors, C-reactive protein was 3.3% (95% CI 1.0–5.6%) higher, interleukin-6 was 2.7% (95% CI 1.1–4.3%) higher, and vitamin D was 11.2% (95% CI 1.0–20.4%) lower. Conclusions Lower outdoor temperature was associated with adverse effects on cholesterol, blood pressure, circulating inflammatory markers, and vitamin D in two older populations. Public health approaches to protect the elderly against low temperatures could help in reducing the levels of several cardiovascular disease risk factors. Biomarkers, outdoor temperature, older adults, cardiovascular disease risk factors Introduction In the UK and most European countries, cardiovascular disease (CVD) risk increases at lower temperatures, a typical element of the cold season.1,2 As CVD risk during the cold season is more markedly increased in older rather than younger adults,3 investigating temperature-related variations in CVD risk factors in older adults is of particular interest. It has been hypothesised that lower outdoor temperatures could exert their adverse effects by increasing the levels of well-established risk factors causally associated with coronary heart disease (CHD),4,5 such as blood pressure6 and lipids.7 However, associations of temperature with recently established causal risk factors for CHD, such as interleukin-6,8 are not well studied.9 Also, low outdoor temperatures may increase the levels of other novel risk factors prospectively associated with CVD (e.g. inflammatory markers, haemostatic markers),10 although the literature supporting this hypothesis is sparse.9,11 Higher outdoor temperature is also a proxy measure for sunlight exposure, and hence potentially related to the level of vitamin D which has consistently been associated with chronic disease incidence although its causal association remains hotly debated.12 Common limitations of previous studies investigating associations of outdoor temperature and CVD risk factors are small sample size,13,14 the specific geographical location,11 and the investigation of clinical populations.15 Therefore, large population-based studies which explore associations of outdoor temperature with a comprehensive range of CVD risk factors are required to improve statistical power and estimate precision. Considering the gaps in knowledge from previous research, the aim of this study was to investigate the strength of relationship between established and novel biological risk factors and outdoor temperature in two large Northern European studies of older adults. Methods and participants Participants from the British Regional Heart Study (BRHS) and the Prospective Study of Pravastatin in the Elderly at Risk (PROSPER) provided informed written consent, which was performed in accordance with the principles of the Declaration of Helsinki. The designs of BRHS and PROSPER, both prospective studies of cardiovascular disease comprising several thousand participants, have been previously described.16 Cardiovascular risk factors measurement (outcomes) For both BRHS and PROSPER, details of measurement values and classification methods for the cardiovascular risk factors were extensively described16 and are briefly reported here in the Supplementary Material, Cardiovascular Risk Factors Measurements. The measurements were carried out during 1997–2000, and the factors included (a) established risk factors, such as systolic and diastolic blood pressure (BP) obtained sitting, and blood lipids (triglycerides, total cholesterol, high density lipoprotein (HDL) cholesterol, and low density lipoprotein (LDL) cholesterol); and (b) novel risk factors, such as inflammatory factors (C-reactive protein (CRP), fibrinogen, interleukin 6 (IL-6)) and plasma viscosity (PV); haemostatic markers (tissue plasminogen activator (t-PA) antigen, fibrin D-dimer, von Willebrand factor (vWF); and vitamin D (VitD). Temperature data National meteorological offices provided daily outdoor mean temperatures for the 24 towns of BRHS and three locations of PROSPER during the study period. Definition of outdoor mean temperature on the examination day (lag 0) has been extensively described elsewhere.16 Statistical methods Descriptive statistics Temperature and the unadjusted outcomes’ levels were examined by month of measurement. Then, excepting total cholesterol, p, HDL-cholesterol, LDL-cholesterol, systolic BP (SBP) and diastolic BP (DBP), all other outcomes were log-transformed for further analysis as their distributions were positively skewed. Associations (main effects) of temperature with the CVD risk factors For log-transformed outcomes, associations were reported as the percentage change in the geometric mean associated with a decrease of 5℃ in mean temperature (5℃ being the standard deviation of daily mean temperature for the years 1997–2000 in the BRHS and PROSPER towns). Associations of temperature with BP variables, HDL-cholesterol, LDL-cholesterol, and total cholesterol, were reported as linear coefficients (absolute change) per decrease of 5℃ in mean temperature. Before being considered for pooling, data from the BRHS and PROSPER were analysed separately due to differences in study design, inclusion criteria and measurement protocols. In BRHS, multilevel linear regression models (level 1 = individual, level 2 = town of examination) were used to take into account clustering within towns.17 Associations were adjusted for established CVD risk factors and possible confounders, such as age, body mass index (BMI), social class, smoking, alcohol consumption, physical activity score and time of day (measurement variable).9 In PROSPER, linear regression models were used to estimate the associations of temperature with the outcomes. Associations were adjusted for the same variables as for BRHS except for physical activity, social class, and time of day which were not ascertained, but for sex and location. In BRHS and PROSPER separately, the proportion of variance associated with temperature from the fully adjusted models was estimated using partial R-squared. We also fitted an interaction between temperature and age (both fitted as continuous variables) to test whether the relationship of temperature with outcomes was particularly marked among older participants. Pooled analysis Regression coefficients from fully adjusted models of BRHS and PROSPER were pooled using a random effects model, to take account of heterogeneity between the two studies where it occurred, for each of the outcomes separately. Sensitivity analysis The cumulative short-term effect of temperature on the CVD risk factors was also investigated, using the temperature moving average of seven days which included lag days from 0–6 prior to the examination day (lag 0–6). An additional adjustment of outdoor temperature (at lag 0) with a seasonal term, such as day length or sine and cosine terms,18 was evaluated. However, since the variance inflation factor scores were between 9–13 for these seasonal terms when included with temperature, collinearity would have been induced and therefore the adjustment was not recommended in this case. Alternatively, we tested an adjustment of temperature with season fitted as binary variable (winter (December–March) vs summer (April–November)). In BRHS, outdoor temperature was also additionally adjusted for indoor temperature (not available in PROSPER). As indoor temperature did not have any effect on the outcomes (all p > 0.05), and did not alter the magnitude of the associations of outdoor temperature with outcomes, it was not considered further. As lung function measurement (forced expiratory volume in one second, or FEV1) was also available in the BRHS, a further sensitivity analysis was carried out adding FEV1 as covariate in models predicting CRP levels, to take into account the possible temperature-related variation in CRP due to poor respiratory health in winter. Results Participants The BRHS and PROSPER participants’ characteristics are shown in Table 1. In BRHS, 4252 men out of 5516 survivors (77%) were examined during the study period. In PROSPER, 5804 participants out of 23,770 (24%) screened individuals participated in the clinical trial of pravastatin vs placebo. PROSPER participants were on average about seven years older than BRHS participants, with a higher percentage of never-smokers, and were less likely to drink alcohol (Table 1). Table 1. The British Regional Heart Study (BRHS) and Prospective Study of Pravastatin in the Elderly at Risk (PROSPER) participant characteristics during examinations (1997–2000). . BRHS men (n = 4252)1 . PROSPER participants (n = 5804)2 . Demographic and background characteristics Sex (male), n (%) 4252 (100) 2806 (48.0) Age (years), mean (SD) 68.7 (5.5) 75.3 (3.3) Social class (manual), n (%) 2166 (51.0) – Physical health Prevalence of stroke/ myocardial infarction, n (%) 370 (8.7) 979 (16.9) Hypertension, n (%) 2703 (63.8) 3592 (61.9) Diabetes, n (%) 380 (9.4) 623 (10.7) BMI, mean (SD) 26.9 (3.7) 26.8 (4.1) Behavioural factors Smoking  Never, n (%) 1233 (29.1) 1969 (33.9)  Ex-smokers, n (%) 2464 (58.0) 2277 (39.2)  Smokers, n (%) 548 (12.9) 1558 (26.8) Alcohol consumption  None, n (%) 431 (10.3) 2576 (44.4)  Occasional/light, n (%) 3 2949 (70.5) 2698 (46.5)  Moderate/heavy, n (%) 4 779 (18.6) 530 (9.1)  Unclassified, n (%) 26 (0.6) – Physical activity (PA) score  Inactive, n (%) 471 (11.5) –  Occasional, n (%) 957 (23.4) –  Light, n (%) 767 (18.7) –  Moderate, n (%) 591 (14.4) –  Moderate vigorous, n (%) 690 (16.8) –  Vigorous, n (%) 621 (15.2) – Biological markers, means (SD)  CRP, mg/l 3.53 (6.86) 5.94 (11.07)  IL-6, pg/ml 3.18 (2.95) 3.40 (3.08)  Fibrinogen, g/l 3.27 (0.74) 3.59 (0.74)  PV, mPa.s 1.285 (0.078) 1.296 (0.077)  t-PA, ng/ml 11.08 (4.44) 11.02 (4.04)  vWF, IU/dl 139.96 (46.19) 140.62 (45.98)  D-dimer, ng/ml 133.58 (210.74) 316.85 (189.48)  Tryglicerides, mmol/l 1.86 (1.08) 1.54 (0.74)  HDL-cholesterol, mmol/l 1.32 (0.34) 1.28 (0.36)  LDL-cholesterol, mmol/l 3.89 (0.97) 3.78 (0.83)  Total cholesterol, mmol/l 6.00 (1.08) 5.66 (0.94)  Vitamin D, ng/ml 20.01 (9.24) 16.57 (9.94)  SBP sitting, mm Hg 149 (24) 155 (22)  DBP sitting, mm Hg 85 (11) 84 (11) . BRHS men (n = 4252)1 . PROSPER participants (n = 5804)2 . Demographic and background characteristics Sex (male), n (%) 4252 (100) 2806 (48.0) Age (years), mean (SD) 68.7 (5.5) 75.3 (3.3) Social class (manual), n (%) 2166 (51.0) – Physical health Prevalence of stroke/ myocardial infarction, n (%) 370 (8.7) 979 (16.9) Hypertension, n (%) 2703 (63.8) 3592 (61.9) Diabetes, n (%) 380 (9.4) 623 (10.7) BMI, mean (SD) 26.9 (3.7) 26.8 (4.1) Behavioural factors Smoking  Never, n (%) 1233 (29.1) 1969 (33.9)  Ex-smokers, n (%) 2464 (58.0) 2277 (39.2)  Smokers, n (%) 548 (12.9) 1558 (26.8) Alcohol consumption  None, n (%) 431 (10.3) 2576 (44.4)  Occasional/light, n (%) 3 2949 (70.5) 2698 (46.5)  Moderate/heavy, n (%) 4 779 (18.6) 530 (9.1)  Unclassified, n (%) 26 (0.6) – Physical activity (PA) score  Inactive, n (%) 471 (11.5) –  Occasional, n (%) 957 (23.4) –  Light, n (%) 767 (18.7) –  Moderate, n (%) 591 (14.4) –  Moderate vigorous, n (%) 690 (16.8) –  Vigorous, n (%) 621 (15.2) – Biological markers, means (SD)  CRP, mg/l 3.53 (6.86) 5.94 (11.07)  IL-6, pg/ml 3.18 (2.95) 3.40 (3.08)  Fibrinogen, g/l 3.27 (0.74) 3.59 (0.74)  PV, mPa.s 1.285 (0.078) 1.296 (0.077)  t-PA, ng/ml 11.08 (4.44) 11.02 (4.04)  vWF, IU/dl 139.96 (46.19) 140.62 (45.98)  D-dimer, ng/ml 133.58 (210.74) 316.85 (189.48)  Tryglicerides, mmol/l 1.86 (1.08) 1.54 (0.74)  HDL-cholesterol, mmol/l 1.32 (0.34) 1.28 (0.36)  LDL-cholesterol, mmol/l 3.89 (0.97) 3.78 (0.83)  Total cholesterol, mmol/l 6.00 (1.08) 5.66 (0.94)  Vitamin D, ng/ml 20.01 (9.24) 16.57 (9.94)  SBP sitting, mm Hg 149 (24) 155 (22)  DBP sitting, mm Hg 85 (11) 84 (11) BMI: body mass index; CRP: C-reactive protein; DBP: diastolic blood pressure; HDL: high density lipoprotein; IL-6: interleukin 6; LDL: low density lipoprotein; PV: plasma viscosity; SBP: systolic blood pressure; SD: standard deviation; t-PA: tissue plasminogen activator; vWF: von Willebrand factor. 1 BRHS men from England and Wales: n = 3804 (89.5%); from Scotland: n = 448 (10.5%). 2 Participants from Glasgow: n = 2520 (43.4%); from Cork: n = 2184 (37.6%), and from Leiden: n = 1100 (19.0%). 3 ≥1 and ≤15 units per week (one unit is approximately one drink, such as one glass of wine). 4 ≥16 units per week (one unit is approximately one drink, such as one glass of wine). Open in new tab Table 1. The British Regional Heart Study (BRHS) and Prospective Study of Pravastatin in the Elderly at Risk (PROSPER) participant characteristics during examinations (1997–2000). . BRHS men (n = 4252)1 . PROSPER participants (n = 5804)2 . Demographic and background characteristics Sex (male), n (%) 4252 (100) 2806 (48.0) Age (years), mean (SD) 68.7 (5.5) 75.3 (3.3) Social class (manual), n (%) 2166 (51.0) – Physical health Prevalence of stroke/ myocardial infarction, n (%) 370 (8.7) 979 (16.9) Hypertension, n (%) 2703 (63.8) 3592 (61.9) Diabetes, n (%) 380 (9.4) 623 (10.7) BMI, mean (SD) 26.9 (3.7) 26.8 (4.1) Behavioural factors Smoking  Never, n (%) 1233 (29.1) 1969 (33.9)  Ex-smokers, n (%) 2464 (58.0) 2277 (39.2)  Smokers, n (%) 548 (12.9) 1558 (26.8) Alcohol consumption  None, n (%) 431 (10.3) 2576 (44.4)  Occasional/light, n (%) 3 2949 (70.5) 2698 (46.5)  Moderate/heavy, n (%) 4 779 (18.6) 530 (9.1)  Unclassified, n (%) 26 (0.6) – Physical activity (PA) score  Inactive, n (%) 471 (11.5) –  Occasional, n (%) 957 (23.4) –  Light, n (%) 767 (18.7) –  Moderate, n (%) 591 (14.4) –  Moderate vigorous, n (%) 690 (16.8) –  Vigorous, n (%) 621 (15.2) – Biological markers, means (SD)  CRP, mg/l 3.53 (6.86) 5.94 (11.07)  IL-6, pg/ml 3.18 (2.95) 3.40 (3.08)  Fibrinogen, g/l 3.27 (0.74) 3.59 (0.74)  PV, mPa.s 1.285 (0.078) 1.296 (0.077)  t-PA, ng/ml 11.08 (4.44) 11.02 (4.04)  vWF, IU/dl 139.96 (46.19) 140.62 (45.98)  D-dimer, ng/ml 133.58 (210.74) 316.85 (189.48)  Tryglicerides, mmol/l 1.86 (1.08) 1.54 (0.74)  HDL-cholesterol, mmol/l 1.32 (0.34) 1.28 (0.36)  LDL-cholesterol, mmol/l 3.89 (0.97) 3.78 (0.83)  Total cholesterol, mmol/l 6.00 (1.08) 5.66 (0.94)  Vitamin D, ng/ml 20.01 (9.24) 16.57 (9.94)  SBP sitting, mm Hg 149 (24) 155 (22)  DBP sitting, mm Hg 85 (11) 84 (11) . BRHS men (n = 4252)1 . PROSPER participants (n = 5804)2 . Demographic and background characteristics Sex (male), n (%) 4252 (100) 2806 (48.0) Age (years), mean (SD) 68.7 (5.5) 75.3 (3.3) Social class (manual), n (%) 2166 (51.0) – Physical health Prevalence of stroke/ myocardial infarction, n (%) 370 (8.7) 979 (16.9) Hypertension, n (%) 2703 (63.8) 3592 (61.9) Diabetes, n (%) 380 (9.4) 623 (10.7) BMI, mean (SD) 26.9 (3.7) 26.8 (4.1) Behavioural factors Smoking  Never, n (%) 1233 (29.1) 1969 (33.9)  Ex-smokers, n (%) 2464 (58.0) 2277 (39.2)  Smokers, n (%) 548 (12.9) 1558 (26.8) Alcohol consumption  None, n (%) 431 (10.3) 2576 (44.4)  Occasional/light, n (%) 3 2949 (70.5) 2698 (46.5)  Moderate/heavy, n (%) 4 779 (18.6) 530 (9.1)  Unclassified, n (%) 26 (0.6) – Physical activity (PA) score  Inactive, n (%) 471 (11.5) –  Occasional, n (%) 957 (23.4) –  Light, n (%) 767 (18.7) –  Moderate, n (%) 591 (14.4) –  Moderate vigorous, n (%) 690 (16.8) –  Vigorous, n (%) 621 (15.2) – Biological markers, means (SD)  CRP, mg/l 3.53 (6.86) 5.94 (11.07)  IL-6, pg/ml 3.18 (2.95) 3.40 (3.08)  Fibrinogen, g/l 3.27 (0.74) 3.59 (0.74)  PV, mPa.s 1.285 (0.078) 1.296 (0.077)  t-PA, ng/ml 11.08 (4.44) 11.02 (4.04)  vWF, IU/dl 139.96 (46.19) 140.62 (45.98)  D-dimer, ng/ml 133.58 (210.74) 316.85 (189.48)  Tryglicerides, mmol/l 1.86 (1.08) 1.54 (0.74)  HDL-cholesterol, mmol/l 1.32 (0.34) 1.28 (0.36)  LDL-cholesterol, mmol/l 3.89 (0.97) 3.78 (0.83)  Total cholesterol, mmol/l 6.00 (1.08) 5.66 (0.94)  Vitamin D, ng/ml 20.01 (9.24) 16.57 (9.94)  SBP sitting, mm Hg 149 (24) 155 (22)  DBP sitting, mm Hg 85 (11) 84 (11) BMI: body mass index; CRP: C-reactive protein; DBP: diastolic blood pressure; HDL: high density lipoprotein; IL-6: interleukin 6; LDL: low density lipoprotein; PV: plasma viscosity; SBP: systolic blood pressure; SD: standard deviation; t-PA: tissue plasminogen activator; vWF: von Willebrand factor. 1 BRHS men from England and Wales: n = 3804 (89.5%); from Scotland: n = 448 (10.5%). 2 Participants from Glasgow: n = 2520 (43.4%); from Cork: n = 2184 (37.6%), and from Leiden: n = 1100 (19.0%). 3 ≥1 and ≤15 units per week (one unit is approximately one drink, such as one glass of wine). 4 ≥16 units per week (one unit is approximately one drink, such as one glass of wine). Open in new tab Outdoor temperature of the day of examination by month In both studies, daily mean temperatures on the day of examination were usually between 4–9℃ from November–April and between 10–16℃ from May–October (see Supplementary Material, eTable 1). CVD risk factors descriptive statistics by month Highest levels of the CVD risk factors analysed were observed from November–April (see Supplementary Material, eTables 2–5). This variation was particularly marked for SBP, DBP, total cholesterol, CRP, IL-6, t-PA, vWF and PV. Conversely, VitD levels were lowest in colder months. Associations of temperature with the CVD risk factors Adjusted associations of mean temperature on day of measurement with the CVD risk factors are shown in Table 2 for each study separately, and pooled. Table 2. The change in the levels of cardiovascular disease (CVD) risk factors for a single standard deviation (5℃) decrease in outdoor mean temperature in the British Regional Heart Study (BRHS) and Prospective Study of Pravastatin in the Elderly at Risk (PROSPER) participants, during examinations (1997–2000). . . . . . POOLED (BRHS + PROSPER)c . . BRHSa . PROSPERb . . Test of heterogeneity . Percentage change (95% CI) . p-Value . Percentage change (95% CI) . p-Value . Percentage change (95% CI) . I2 (%) . p-Value . CRP, mg/l 4.1 (0.7–7.3) 0.017 2.4 (–0.8–5.7) 0.075 3.3 (1.0–5.6) 0.0 0.468 IL-6, pg/ml 1.8 (–1.3–4.8) 0.246 3.0 (1.1–4.9) 0.001 2.7 (1.1–4.3) 0.0 0.525 Fibrinogen, g/l 0.5 (–0.5–1.6) 0.286 0.8 (0.2–1.4) 0.007 0.7 (0.2–1.3) 0.0 0.677 t-PA, ng/ml 2.5 (0.6–4.4) 0.010 1.7 (0.6–2.8) <0.001 1.9 (1.0–2.9) 40.6 0.461 PV, mPa.s 0.4 (0.2–0.6) <0.001 0.4 (0.3–0.6) <0.001 0.4 (0.3–0.5) 0.0 1.000 vWF, IU/dl –1.0 (–2.6–0.7) 0.268 1.0 (0.0–2.0) 0.029 0.1 (–1.7–2.1) 74.0 0.050 D-dimer, ng/ml 1.6 (–1.5–4.6) 0.292 –0.6 (–2.1–0.9) 0.482 0.1 (–1.9–2.2) 0.0 0.516 Vitamin D, ng/ml –6.1 (–9.1– –3.2) <0.001 –16.0 (–17.5– –14.5) <0.001 –11.2 (–20.4– –1.0) 97.3 <0.001 Triglycerides, mmol/l 1.5 (–0.7–3.6) 0.175 –0.1 (–1.3–1.1) 0.442 0.4 (–1.0–1.9) 39.3 0.199 Absolute change (95% CI) p-Value Absolute change (95% CI) p-Value Absolute change (95% CI) I2 (%) p-Value HDL-cholesterol, mmol/l 0.00 (–0.01–0.02) 0.844 0.03 (0.02–0.04) <0.001 0.02 (–0.01–0.05) 90.6 0.001 LDL-cholesterol, mmol/l 0.05 (0.00–0.09) 0.039 0.02 (0.00–0.05) 0.038 0.03 (0.01–0.05) 0.0 0.346 Total cholesterol, mmol/l 0.06 (0.01–0.11) 0.015 0.04 (0.01–0.06) 0.004 0.04 (0.02–0.07) 0.0 0.464 SBP sitting, mm Hg 1.22 (0.29–2.16) 0.010 1.08 (0.45–1.71) <0.001 1.12 (0.60–1.64) 0.0 0.796 DBP sitting, mm Hg 0.47 (0.04–0.90) 0.032 0.13 (–0.20–0.46) 0.270 0.27 (–0.06–0.60) 34.5 0.217 . . . . . POOLED (BRHS + PROSPER)c . . BRHSa . PROSPERb . . Test of heterogeneity . Percentage change (95% CI) . p-Value . Percentage change (95% CI) . p-Value . Percentage change (95% CI) . I2 (%) . p-Value . CRP, mg/l 4.1 (0.7–7.3) 0.017 2.4 (–0.8–5.7) 0.075 3.3 (1.0–5.6) 0.0 0.468 IL-6, pg/ml 1.8 (–1.3–4.8) 0.246 3.0 (1.1–4.9) 0.001 2.7 (1.1–4.3) 0.0 0.525 Fibrinogen, g/l 0.5 (–0.5–1.6) 0.286 0.8 (0.2–1.4) 0.007 0.7 (0.2–1.3) 0.0 0.677 t-PA, ng/ml 2.5 (0.6–4.4) 0.010 1.7 (0.6–2.8) <0.001 1.9 (1.0–2.9) 40.6 0.461 PV, mPa.s 0.4 (0.2–0.6) <0.001 0.4 (0.3–0.6) <0.001 0.4 (0.3–0.5) 0.0 1.000 vWF, IU/dl –1.0 (–2.6–0.7) 0.268 1.0 (0.0–2.0) 0.029 0.1 (–1.7–2.1) 74.0 0.050 D-dimer, ng/ml 1.6 (–1.5–4.6) 0.292 –0.6 (–2.1–0.9) 0.482 0.1 (–1.9–2.2) 0.0 0.516 Vitamin D, ng/ml –6.1 (–9.1– –3.2) <0.001 –16.0 (–17.5– –14.5) <0.001 –11.2 (–20.4– –1.0) 97.3 <0.001 Triglycerides, mmol/l 1.5 (–0.7–3.6) 0.175 –0.1 (–1.3–1.1) 0.442 0.4 (–1.0–1.9) 39.3 0.199 Absolute change (95% CI) p-Value Absolute change (95% CI) p-Value Absolute change (95% CI) I2 (%) p-Value HDL-cholesterol, mmol/l 0.00 (–0.01–0.02) 0.844 0.03 (0.02–0.04) <0.001 0.02 (–0.01–0.05) 90.6 0.001 LDL-cholesterol, mmol/l 0.05 (0.00–0.09) 0.039 0.02 (0.00–0.05) 0.038 0.03 (0.01–0.05) 0.0 0.346 Total cholesterol, mmol/l 0.06 (0.01–0.11) 0.015 0.04 (0.01–0.06) 0.004 0.04 (0.02–0.07) 0.0 0.464 SBP sitting, mm Hg 1.22 (0.29–2.16) 0.010 1.08 (0.45–1.71) <0.001 1.12 (0.60–1.64) 0.0 0.796 DBP sitting, mm Hg 0.47 (0.04–0.90) 0.032 0.13 (–0.20–0.46) 0.270 0.27 (–0.06–0.60) 34.5 0.217 BMI: body mass index; CI: confidence interval; CRP: C-reactive protein; DBP: diastolic blood pressure; HDL: high density lipoprotein; IL-6: interleukin 6; LDL: low density lipoprotein; PV: plasma viscosity; SBP: systolic blood pressure; t-PA: tissue plasminogen activator; vWF: von Willebrand factor. a Multilevel linear regression models (level 1 = individual, level 2 = town of examination) were used. The models were adjusted for age, social class, BMI, smoking, alcohol consumption, physical activity, and time of measurement. Complete case analysis (n = 3832) b Linear regression models were used. The models were adjusted for town, age, BMI, smoking, alcohol consumption, and sex. Complete case analysis (n = 5804) c Results from the two studies were pooled using a random effects model. The command metan with the option random available in Stata/SE 14. The percentage of variation across studies that is due to heterogeneity was reported using the I2 statistic. Open in new tab Table 2. The change in the levels of cardiovascular disease (CVD) risk factors for a single standard deviation (5℃) decrease in outdoor mean temperature in the British Regional Heart Study (BRHS) and Prospective Study of Pravastatin in the Elderly at Risk (PROSPER) participants, during examinations (1997–2000). . . . . . POOLED (BRHS + PROSPER)c . . BRHSa . PROSPERb . . Test of heterogeneity . Percentage change (95% CI) . p-Value . Percentage change (95% CI) . p-Value . Percentage change (95% CI) . I2 (%) . p-Value . CRP, mg/l 4.1 (0.7–7.3) 0.017 2.4 (–0.8–5.7) 0.075 3.3 (1.0–5.6) 0.0 0.468 IL-6, pg/ml 1.8 (–1.3–4.8) 0.246 3.0 (1.1–4.9) 0.001 2.7 (1.1–4.3) 0.0 0.525 Fibrinogen, g/l 0.5 (–0.5–1.6) 0.286 0.8 (0.2–1.4) 0.007 0.7 (0.2–1.3) 0.0 0.677 t-PA, ng/ml 2.5 (0.6–4.4) 0.010 1.7 (0.6–2.8) <0.001 1.9 (1.0–2.9) 40.6 0.461 PV, mPa.s 0.4 (0.2–0.6) <0.001 0.4 (0.3–0.6) <0.001 0.4 (0.3–0.5) 0.0 1.000 vWF, IU/dl –1.0 (–2.6–0.7) 0.268 1.0 (0.0–2.0) 0.029 0.1 (–1.7–2.1) 74.0 0.050 D-dimer, ng/ml 1.6 (–1.5–4.6) 0.292 –0.6 (–2.1–0.9) 0.482 0.1 (–1.9–2.2) 0.0 0.516 Vitamin D, ng/ml –6.1 (–9.1– –3.2) <0.001 –16.0 (–17.5– –14.5) <0.001 –11.2 (–20.4– –1.0) 97.3 <0.001 Triglycerides, mmol/l 1.5 (–0.7–3.6) 0.175 –0.1 (–1.3–1.1) 0.442 0.4 (–1.0–1.9) 39.3 0.199 Absolute change (95% CI) p-Value Absolute change (95% CI) p-Value Absolute change (95% CI) I2 (%) p-Value HDL-cholesterol, mmol/l 0.00 (–0.01–0.02) 0.844 0.03 (0.02–0.04) <0.001 0.02 (–0.01–0.05) 90.6 0.001 LDL-cholesterol, mmol/l 0.05 (0.00–0.09) 0.039 0.02 (0.00–0.05) 0.038 0.03 (0.01–0.05) 0.0 0.346 Total cholesterol, mmol/l 0.06 (0.01–0.11) 0.015 0.04 (0.01–0.06) 0.004 0.04 (0.02–0.07) 0.0 0.464 SBP sitting, mm Hg 1.22 (0.29–2.16) 0.010 1.08 (0.45–1.71) <0.001 1.12 (0.60–1.64) 0.0 0.796 DBP sitting, mm Hg 0.47 (0.04–0.90) 0.032 0.13 (–0.20–0.46) 0.270 0.27 (–0.06–0.60) 34.5 0.217 . . . . . POOLED (BRHS + PROSPER)c . . BRHSa . PROSPERb . . Test of heterogeneity . Percentage change (95% CI) . p-Value . Percentage change (95% CI) . p-Value . Percentage change (95% CI) . I2 (%) . p-Value . CRP, mg/l 4.1 (0.7–7.3) 0.017 2.4 (–0.8–5.7) 0.075 3.3 (1.0–5.6) 0.0 0.468 IL-6, pg/ml 1.8 (–1.3–4.8) 0.246 3.0 (1.1–4.9) 0.001 2.7 (1.1–4.3) 0.0 0.525 Fibrinogen, g/l 0.5 (–0.5–1.6) 0.286 0.8 (0.2–1.4) 0.007 0.7 (0.2–1.3) 0.0 0.677 t-PA, ng/ml 2.5 (0.6–4.4) 0.010 1.7 (0.6–2.8) <0.001 1.9 (1.0–2.9) 40.6 0.461 PV, mPa.s 0.4 (0.2–0.6) <0.001 0.4 (0.3–0.6) <0.001 0.4 (0.3–0.5) 0.0 1.000 vWF, IU/dl –1.0 (–2.6–0.7) 0.268 1.0 (0.0–2.0) 0.029 0.1 (–1.7–2.1) 74.0 0.050 D-dimer, ng/ml 1.6 (–1.5–4.6) 0.292 –0.6 (–2.1–0.9) 0.482 0.1 (–1.9–2.2) 0.0 0.516 Vitamin D, ng/ml –6.1 (–9.1– –3.2) <0.001 –16.0 (–17.5– –14.5) <0.001 –11.2 (–20.4– –1.0) 97.3 <0.001 Triglycerides, mmol/l 1.5 (–0.7–3.6) 0.175 –0.1 (–1.3–1.1) 0.442 0.4 (–1.0–1.9) 39.3 0.199 Absolute change (95% CI) p-Value Absolute change (95% CI) p-Value Absolute change (95% CI) I2 (%) p-Value HDL-cholesterol, mmol/l 0.00 (–0.01–0.02) 0.844 0.03 (0.02–0.04) <0.001 0.02 (–0.01–0.05) 90.6 0.001 LDL-cholesterol, mmol/l 0.05 (0.00–0.09) 0.039 0.02 (0.00–0.05) 0.038 0.03 (0.01–0.05) 0.0 0.346 Total cholesterol, mmol/l 0.06 (0.01–0.11) 0.015 0.04 (0.01–0.06) 0.004 0.04 (0.02–0.07) 0.0 0.464 SBP sitting, mm Hg 1.22 (0.29–2.16) 0.010 1.08 (0.45–1.71) <0.001 1.12 (0.60–1.64) 0.0 0.796 DBP sitting, mm Hg 0.47 (0.04–0.90) 0.032 0.13 (–0.20–0.46) 0.270 0.27 (–0.06–0.60) 34.5 0.217 BMI: body mass index; CI: confidence interval; CRP: C-reactive protein; DBP: diastolic blood pressure; HDL: high density lipoprotein; IL-6: interleukin 6; LDL: low density lipoprotein; PV: plasma viscosity; SBP: systolic blood pressure; t-PA: tissue plasminogen activator; vWF: von Willebrand factor. a Multilevel linear regression models (level 1 = individual, level 2 = town of examination) were used. The models were adjusted for age, social class, BMI, smoking, alcohol consumption, physical activity, and time of measurement. Complete case analysis (n = 3832) b Linear regression models were used. The models were adjusted for town, age, BMI, smoking, alcohol consumption, and sex. Complete case analysis (n = 5804) c Results from the two studies were pooled using a random effects model. The command metan with the option random available in Stata/SE 14. The percentage of variation across studies that is due to heterogeneity was reported using the I2 statistic. Open in new tab Pooled estimates showed that with a 5℃ lower mean temperature, total cholesterol was 0.04 mmol/l (95% confidence interval (CI) 0.02–0.07) higher, LDL cholesterol was 0.02 mmol/l (95% CI 0.01–0.05) higher, and SBP was 1.12 mm Hg (95% CI 0.60–1.64) higher. Among novel CVD risk factors, CRP was 3.3% (95% CI 1.0–5.6%) higher, IL-6 was 2.7% (95% CI 1.1–4.3%) higher, t-PA was 1.9% (95% CI 1.0–2.9%) higher, fibrinogen was 0.7% (95% CI 0.2–1.3%) higher, and plasma viscosity was 0.4% (95% CI 0.3–0.5%) higher. There was no evidence of heterogeneity between studies (p-values > 0.05). With a 5℃ lower mean temperature, VitD was 11.2% (95% CI 1.0–20.4%) lower. In this case, there was evidence of heterogeneity between studies (I2 = 97.3%; p-value < 0.001), though the effect was in the same direction and statistically significant for both studies. Associations of temperature with DBP, vWF and D-dimer, triglycerides, and HDL-cholesterol were not statistically significant. Results for HDL-cholesterol suggested heterogeneity (I2 = 90.6%; p-value = 0.001) with association of a decrease in temperature significant for the PROSPER study only. Proportion of variance in risk factors explained by temperature The highest proportion of variance was observed when the outcome analysed was VitD (5.1% and 5.6% in the BRHS and PROSPER fully adjusted models respectively). In each of the models, and other outcomes analysed, the proportion of variance associated with mean temperature was less than 1% (Supplementary Material, eTable 6). Interactions between temperature and age Interaction effects of temperature with age on the outcomes levels were mainly not significant (data not shown). However, interactions were found in PROSPER alone for VitD and HDL-cholesterol. A 5℃ decrease in mean temperature was associated with an additional decrease of –0.8% per year of age (95% CI –1.4– –0.3%) for Vitamin D, and +0.003 mmol/l per year of age (95% CI 0–0.006) for HDL-cholesterol. No interactions were found in BRHS. Sensitivity analysis Cumulative short-term associations of temperature up to one week (lag 0–6) prior to the examination day with the CVD risk factors levels were observed (not shown). As the magnitude of the associations was very similar to associations using temperature at lag 0 (primary analysis), only associations at lag 0 were presented. An additional adjustment for season fitted as binary variable (winter vs summer) barely changed the magnitude of the associations of outdoor temperature (results were not shown). In BRHS, an additional adjustment for lung function (FEV1) did not substantially change the effect of CRP: the percentage increase in CRP due to a decrease in temperature was 4.1% (0.7–7.3%) and 4.6% (1.4–7.8%) for models without and with lung function. Discussion To our knowledge, the pooled analysis of the BRHS and PROSPER is the largest investigation of the relationships between outdoor temperature and an extensive range of CVD risk factors, both established and novel in older European people. The CVD risk factors investigated here were selected for two reasons: first, there was published evidence of seasonal variation, with higher levels observed in the cold season (November–April); second, there was published evidence of independent association with CVD events in meta-analyses of prospective population-based studies.19–22 Overall findings Lower outdoor temperature, measured on the day of clinical examination, was associated with higher levels of most CVD risk factors analysed. Conversely, lower outdoor temperature was associated with a lower VitD. The direction and magnitude of these associations were similar in comparison with other studies,6,7,11 and persisted after adjustment for classic risk factors such as age, BMI, smoking, alcohol consumption, and physical activity. The findings were similar when using the outdoor temperature moving average of seven days, which included lag days from 0–6 prior to the examination day (lag 0–6). In fully adjusted models, the proportion of variance in risk factors explained by temperature was much smaller than other risk factors, being around 1% of the total variance (except for VitD, where variance explained was approximately 5%). There was no consistent evidence of an interaction of temperature with age on the wide range of CVD risk factors analysed. These findings would be consistent with the suggestions from previous studies that, in addition to established risk factors such as cholesterol7 and BP,6 circulating inflammatory markers,9 and VitD23 showed strong associations with outdoor temperature and may contribute to increased incidence of CVD in winter.1 The association of temperature with SBP, LDL-cholesterol and IL-6 levels may be particularly relevant, as previous trials and Mendelian randomization (MR) studies support their causal role in CHD risk.4,5,24 Established CVD risk factors In this study lower outdoor temperatures were significantly associated with higher levels of SBP consistently with previous findings.25 The association with DBP was weaker and non-significant. Seasonal variation in SBP was previously shown to be greater in older than in younger subjects (while DBP was similar), and highly significantly related to outdoor temperature.26 We found decrease in temperature was associated with increased total cholesterol and LDL-cholesterol, as previously reported.27 In our study, a decrease of about 10℃ in temperatures would be associated with an increase of 0.06 mmol/l in LDL-cholesterol. According to previous studies, this absolute increase in LDL-cholesterol leads to an increase of approximately 1% in CVD mortality risk.4 The importance of HDL-cholesterol as a marker of CHD risk has been emphasised through its inclusion in the Framingham Risk Score.28 When pooling results from the two studies, we found no clear association between temperature and HDL-cholesterol although a positive association was seen in PROSPER. Lastly, associations of temperature with triglycerides were not significant as observed in previous studies.7 Novel CVD risk factors A decrease in temperature was associated with increased circulating levels of markers of inflammation, such as IL-6, CRP, fibrinogen and plasma viscosity. The inflammatory hypothesis of CVD is currently being formally tested in randomized controlled trials (RCTs).24 To date, MR studies for IL-6 suggested a causal role in CHD, in contrast to null associations in MR studies for CRP and fibrinogen.8 Therefore, the findings on IL-6 are particularly important: in this study a decrease of about 10℃ in temperatures (difference between the coldest and warmest month, January–August) would be associated with an increase of 0.06 pg/ml in IL-6 levels. According to previous IL-6 observational data this absolute difference was associated to an increase of 4.5% in CVD deaths.29 This broad estimation is in line with previous studies which took place in the same years (1998–2007) and attributed to temperatures 7% of the winter mortality in England and Wales.30 It is also possible that an acute (or short-term) effect of outdoor temperature may be more marked on rapidly responding CVD risk factors, such as CRP.31 The CRP behaviour may explain why it provides closer associations and better predictions of CVD events in the short-term than other markers of inflammation. The associations of temperature with other specific markers of inflammation we studied, such as fibrinogen and plasma viscosity, were smaller in comparison with CRP, as previously reported.15 Findings for PV and t-PA are similar in comparison with previous studies which observed higher levels of these factors in winter,9 although the effect of temperature was not specifically tested. To our knowledge these associations with temperature are novel, and have not been previously published. On the other hand, the association of temperature with vWF and fibrin D-dimer was not significant. The seasonal variation in temperature did not show a good agreement with variations observed in vWF and D-dimer: vWF’s seasonal peak was previously observed in early spring (between March–May)9 when outdoor temperature already started its annual average increase from February; D-dimer seemed to have an unusual seasonal variation, with peaks in February/March and August/September.32 VitD Findings for VitD showed strong associations with temperature in pooled analysis, though this varied between the studies. However, for VitD specifically, temperature is likely to be a proxy of exposure to sunlight, which is the real determinant. In our study a decrease of about 10℃ in temperatures would be associated with a decrease of approximately 4 ng/ml (=10 nmol/l) in VitD levels. According to previous observational studies, this absolute decrease in VitD is associated with an increase of approximately 4% in CVD deaths and events12 although any causal role remains contentious. Strengths and limitations By pooling PROSPER and BRHS we substantially improved statistical power and precision in comparison with findings reported in other studies of older adults. The participants lived mostly in the UK but also in Ireland and the Netherlands. The PROSPER study included both women and men. Moreover, novel CVD risk factors measurements in both studies were performed over the same time period, in the same Glasgow University laboratories using the same assays. However, the two study designs are different and this may partially explain the heterogeneity of the findings for HDL cholesterol, as well as the interactions of temperature with age; the PROSPER participants in comparison with the BRHS participants were about seven years older on average, with a higher percentage of never-smokers, and less likely to drink alcohol. They were also at elevated CVD risk and around half had prevalent CVD. Due to the nature of our data and risk of collinearity between temperature and other seasonal terms, it was not possible to distinguish between temperature-related effects and effects due to other factors which are known to vary by season: for example, in winter higher prevalence of influenza or other respiratory viruses or diseases, such as rheumatic disorders, may be relevant to the CRP seasonal variation. Despite this limitation, we took into account of season as binary variable (winter vs summer), and alternatively fitting respiratory health in sensitivity analysis: an additional adjustment for lung function was performed and specifically when using CRP as outcome. The results still showed that lower outdoor temperatures were significantly associated with an increase in the outcome levels. Moreover, although indoor temperature was not available in the PROSPER, we added this variable in the BRHS models and we showed the effect of outdoor temperatures was not confounded by indoor temperatures. Implications Our study provides robust evidence that outdoor temperature is associated with variations in the major CVD risk factors in older adults. This study increased generalisability of existing evidence from northern European older populations and is consistent with the hypothesis that inflammation markers, on top of BP and LDL-cholesterol changes, could play a key role in intermediate processes leading to the cold-related CVD mortality. Also, there was no consistent evidence of an interaction of temperature with age (participants in the two studies ranged from 60–82 years old) on the wide range of CVD risk factors analysed; this finding suggested that the effect of low temperature on CVD risk may apply to the full age range of older adults. Public health approaches to protect elderly populations against low temperatures could help in reducing levels of several CVD risk factors, and thus CVD risk itself, in winter. Conclusions Variations of outdoor temperature in the short-term were associated with variations in the majority of CVD risk factors analysed. Associations were strongest with inflammatory factors (particularly CRP, and its major cytokine driver, IL-6) and VitD, followed by associations with SBP, and cholesterol variables. Better protection against low temperatures could help in reducing the levels of several CVD risk factors. Author contribution Study, concept, design, and acquisition of data: PHW, SGW, RWM; data collection: LL; laboratory analysis: PW; analysis and interpretation of data: CS, SJEB, RWM; drafting the article: CS, RWM; revising the article critically for important intellectual content: CS, SJEB, PHW, SGW, GDOL, BJJ, IF, NS, RWM; final approval of the version to be published: CS, SJEB, PHW, SGW, GDOL, BJJ, LL, PW, IF, NS, RWM. 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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by a British Heart Foundation (BHF) project grant (PG/13/41/30304) which supported CS. The BRHS is supported by a BHF programme grant (RG/13/16/30528). The funders had no role in the design, methods, subject recruitment, data collections, analysis or preparation of the paper. The views expressed in this publication are those of the author(s) and not necessarily those of the funders. References 1 Fowler T , Southgate RJ, Waite Tet al. Excess winter deaths in Europe: A multi-country descriptive analysis . Eur J Public Health 2014 ; 25 : 339 – 345 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Abrignani MG , Corrao S, Biondo GBet al. Effects of ambient temperature, humidity, and other meteorological variables on hospital admissions for angina pectoris . Eur J Prev Cardiol 2012 ; 19 : 342 – 348 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Office of National Statistics. Excess winter mortality in England and Wales: 2014, 2013–14 and 2012–13, https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/bulletins/excesswintermortalityinenglandandwales/2014-11-28 . 4 Cholesterol Treatment Trialists Collaboration . Efficacy and safety of more intensive lowering of LDL cholesterol: A meta-analysis of data from 170 000 participants in 26 randomised trials . Lancet 2010 ; 376 : 1670 – 1681 . Crossref Search ADS PubMed WorldCat 5 Neal B , MacMahon S, Chapman N. Effects of ACE inhibitors, calcium antagonists, and other blood-pressure-lowering drugs: Results of prospectively designed overviews of randomised trials. Blood Pressure Lowering Treatment Trialists' Collaboration . Lancet 2000 ; 356 : 1955 – 1964 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Halonen JI , Zanobetti A, Sparrow Det al. Relationship between Outdoor Temperature and Blood Pressure . Occup Environ Med 2011 ; 68 : 296 – 301 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Halonen JI , Zanobetti A, Sparrow Det al. Outdoor temperature is associated with serum HDL and LDL . Environ Res 2011 ; 111 : 281 – 287 . Google Scholar Crossref Search ADS PubMed WorldCat 8 The Interleukin-6 Receptor Mendelian Randomisation Analysis (IL6R MR) Consortium . The interleukin-6 receptor as a target for prevention of coronary heart disease: A Mendelian randomisation analysis . Lancet 2012 ; 379 : 1214 – 1224 . Crossref Search ADS PubMed WorldCat 9 Rudnicka AR , Rumley A, Lowe GDet al. Diurnal, seasonal, and blood-processing patterns in levels of circulating fibrinogen, fibrin D-dimer, C-reactive protein, tissue plasminogen activator, and von Willebrand factor in a 45-year-old population . Circulation 2007 ; 115 : 996 – 1003 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Peters SA , Woodward M, Rumley Aet al. Plasma and blood viscosity in the prediction of cardiovascular disease and mortality in the Scottish Heart Health Extended Cohort Study . Eur J Prev Cardiol 2016 ; 0 : 1 – 7 . Google Scholar OpenURL Placeholder Text WorldCat 11 Halonen JI , Zanobetti A, Sparrow Det al. Associations between outdoor temperature and markers of inflammation: A cohort study . Environ Health 2010 ; 9 : 42 – 42 , doi:10.1186/1476-069X-9-42. Google Scholar Crossref Search ADS PubMed WorldCat 12 Khaw K-T , Luben R, Wareham N. Serum 25-hydroxyvitamin D, mortality, and incident cardiovascular disease, respiratory disease, cancers, and fractures: A 13-y prospective population study . Am J Clin Nutr 2014 ; 100 : 1361 – 1370 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Stout RW , Crawford V. Seasonal variations in fibrinogen concentrations among elderly people . Lancet 1991 ; 338 : 9 – 13 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Otto C , Donner MG, Schwandt Pet al. Seasonal variations of hemorheological and lipid parameters in middle-aged healthy subjects . Clinica Chimica Acta 1996 ; 256 : 87 – 94 . Google Scholar Crossref Search ADS WorldCat 15 Schneider A , Panagiotakos D, Picciotto Set al. Air temperature and inflammatory responses in myocardial infarction survivors . Epidemiology 2008 ; 19 : 391 – 400 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Sartini C , Barry SJE, Wannamethee SGet al. Effect of cold spells and their modifiers on cardiovascular disease events: Evidence from two prospective studies . Int J Cardiol 2016 ; 18 : 275 – 283 . Google Scholar Crossref Search ADS WorldCat 17 Morris RW , Whincup PH, Emberson JRet al. North-south gradients in Britain for stroke and CHD: Are they explained by the same factors? Stroke 2003 ; 34 : 2604 – 2609 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Stolwijk AM , Straatman H, Zielhuis GA. Studying seasonality by using sine and cosine functions in regression analysis . J Epidemiol Community Health 1999 ; 53 : 235 – 238 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Danesh J , Collins R, Peto Ret al. Haematocrit, viscosity, erythrocyte sedimentation rate: Meta-analyses of prospective studies of coronary heart disease . Eur Heart J 2000 ; 21 : 515 – 520 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Prospective Studies Collaboration . Age-specific relevance of usual blood pressure to vascular mortality: A meta-analysis of individual data for one million adults in 61 prospective studies . Lancet 2002 ; 360 : 1903 – 1913 . Crossref Search ADS PubMed WorldCat 21 Di Angelantonio E , Gao P, Pennells Let al. Lipid-related markers and cardiovascular disease prediction . JAMA 2012 ; 307 : 2499 – 2506 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 22 Kaptoge S , Seshasai SR, Gao Pet al. Inflammatory cytokines and risk of coronary heart disease: New prospective study and updated meta-analysis . Eur Heart J 2014 ; 35 : 578 – 589 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Berry DJ , Hesketh K, Power Cet al. Vitamin D status has a linear association with seasonal infections and lung function in British adults . Br J Nutr 2011 ; 106 : 1433 – 1440 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Ridker PM . Closing the loop on inflammation and atherothrombosis: Why perform the CIRT and CANTOS trials? Trans Am Clin Climatol Assoc 2013 ; 124 : 174 – 190 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 25 Alperovitch A , Lacombe JM, Hanon Oet al. Relationship between blood pressure and outdoor temperature in a large sample of elderly individuals: The Three-City study . Arch Intern Med 2009 ; 169 : 75 – 80 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Brennan PJ , Greenberg G, Miall WEet al. Seasonal variation in arterial blood pressure . Br Med J 1982 ; 285 : 919 – 923 . Google Scholar Crossref Search ADS WorldCat 27 Hong YC , Kim H, Oh SYet al. Association of cold ambient temperature and cardiovascular markers . Sci Total Environ 2012 ; vol. 435–436 : 74 – 79 . Google Scholar Crossref Search ADS WorldCat 28 Wilson PWF , D’Agostino RB, Levy Det al. Prediction of coronary heart disease using risk factor categories . Circulation 1998 ; 97 : 1837 – 1847 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Sattar N , Murray HM, Welsh Pet al. Are markers of inflammation more strongly associated with risk for fatal than for nonfatal vascular events? PLoS Med 2009 ; 6 : e1000099 – e1000099 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Brown G , Fearn V, Wells C. Exploratory analysis of seasonal mortality in England and Wales, 1998 to 2007 . Health Stat Q 2010 ; 48 : 58 – 80 . Google Scholar Crossref Search ADS WorldCat 31 Pepys MB , Hirschfield GM. C-reactive protein: A critical update . J Clin Invest 2003 ; 111 : 1805 – 1812 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Berry DJ , Hypponen E, Cortina-Borja M. Investigating the association of vitamin D seasonality on inflammatory and hemostatic markers . Chronobiol Int 2013 ; 30 : 786 – 795 . Google Scholar Crossref Search ADS PubMed WorldCat © The European Society of Cardiology 2017 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 2017

Abstract Background It has been previously shown in patients with heart failure that exercise-based rehabilitation programmes may improve functional capacity and autonomic response. The aim of this study was to investigate this issue further by evaluating whether an association exists between autonomic adaptations and improvements of aerobic capacity in a general population of coronary artery disease patients undergoing cardiac rehabilitation. Methods Ninety consecutive patients (age 60 ± 11 years) attended a rehabilitation programme of moderate continuous training (25 ± 8 sessions, 2–3 sessions/week). Functional capacity expressed as oxygen uptake (peak VO2) and autonomic function expressed as chronotropic response and heart rate recovery were evaluated by cardiopulmonary exercise tests before and after the rehabilitation programme. According to the expected mean increase in functional capacity, coronary artery disease patients were divided into two groups: those who improved peak VO2 by more than 2.6 ml/kg/min (R group) and those who did not (NR group). Effects of the rehabilitation programme were compared in R and NR groups. Results The number and intensity of exercise sessions did not differ between R (N = 39) and NR (N = 51) groups. However, only R patients improved chronotropic response (R: from 45.1 ± 16.9% to 72.7 ± 34.1%, P < 0.01; NR: from 49.3 ± 18.6% to 48.2 ± 36.5%, P = NS) and heart rate recovery (R: from 16.9 ± 7.0 bpm to 21.0 ± 8.7 bpm, P < 0.01; NR: from 15.2 ± 9.9 bpm to 15.8 ± 8.5 bpm, P = NS). After training both chronotropic response and heart rate recovery were significantly higher in R than NR patients. Conclusions The improvement in aerobic capacity of coronary artery disease patients following exercise-based cardiac rehabilitation programmes is associated with positive adaptations of autonomic function. Coronary artery disease, functional capacity, heart rate recovery, chronotropic response, cardiopulmonary exercise testing, cardiac rehabilitation Introduction There is convincing evidence on the benefits of both moderate continuous training and aerobic interval training in patients with cardiovascular disease.1,2 More specifically, exercise training is a safe therapy able to improve functional capacity, expressed as peak exercise oxygen consumption (peak VO2), which represents the most powerful predictor of cardiac and overall mortality.3–5 Exercise training also improves autonomic function6,7 as demonstrated by positive adaptations of sympatho-vagal balance markers, such as chronotropic response to exercise (CR) and heart rate recovery after exercise (HRR). CR appears to reflect a combination of parasympathetic withdrawal, sympathetic activation and good sensitivity of the beta adrenergic receptor pathway in sinoatrial tissue while HRR seems mainly to reflect parasympathetic reactivation.8–10 Besides functional capacity, both CR and HRR are independent prognostic markers of major adverse cardiovascular events and overall mortality.11–13 As changes in aerobic capacity and autonomic function after exercise training have generally been examined separately in various subjects populations,14,15 the purpose of this study was to evaluate whether an association exists between improvements in aerobic capacity and autonomic indices of fitness after exercise training. This would allow elucidating the role of cardiac autonomic modulations in determining different responses of functional capacity to a moderate continuous training in patients with established coronary artery disease (CAD). Methods Approval was obtained from our ethical committee on human research and the patients gave their written informed consent. From October 2011 to December 2014, 90 consecutive CAD patients underwent a cardiac rehabilitation programme (CRP) at the Cardiovascular Prevention and Rehabilitation Unit of Don Gnocchi Foundation, University of Parma. The criteria for inclusion were established ischaemic heart disease (stable CAD or acute coronary syndrome with ST elevation or non-ST elevation myocardial infarction) without clinical evidence of congestive heart failure (CHF) treated with coronary artery bypass surgery, with percutaneous coronary intervention or with medical therapy and the presence of sinus rhythm at admission within 30 days after hospital discharge. Laboratory markers were obtained at admission of CRP; 12-lead ECG, two-dimensional (2D) echocardiography and cardiopulmonary exercise test (C-PET) were performed at admission and discharge from CRP. Cardiopulmonary exercise test Incremental and maximal cardiopulmonary tests were performed with the Cosmed Quark C-PET system until maximal perceived exertion, assessed using the Borg scale.16 Further criteria for termination of C-PET were determined according to the scientific statement from the American Heart Association on exercise standards for testing and training.17 Patients pedalled in the upright position on an electronically braked cycle ergometer at a constant rate of 60 rpm after an acclimatisation period of 3 minutes. The protocol included a warm-up period (5 Watt load) for 3 minutes, an incremental ramp protocol based on predicted maximal functional capacity to achieve in 8–12 minutes, and an unloading recovery of 3 minutes. Volumes and gases (oxygen and carbon dioxide) were calibrated before each test. Functional capacity was evaluated as peak oxygen uptake (VO2) (ml/kg/min) (mean VO2 over the last 30 seconds of exercise), and as a percentage of the maximum theoretical VO2 prediction value estimated for the corresponding age, gender, weight and height (VO2%). The slope of the regression between ventilation (VE) and carbon dioxide production (VCO2) (VE/VCO2 slope) and between VO2 and exercise load (VO2/Watt slope), and the ratios between VE and, respectively, VO2 and VCO2 at peak values (VE/VO2 peak and VE/VCO2 peak) were also derived during the incremental exercise. The ventilatory threshold was measured by the V-slope method. Systolic and diastolic arterial pressures were measured with an arm cuff at rest and at exercise peak. The value for the recovery of heart rate was defined as the reduction in the heart rate from the rate at peak exercise to the rate one minute after the cessation of exercise.18 CR was evaluated as the percentage of chronotropic reserve, i.e. Brubaker and Kitzman.8 CR=peakHR-restHRapmHR-restHR×100 (1) with apm HR the age-predicted maximal heart rate (HR), in bpm, estimated by Tanaka’s regression formula:8 apmHR=208-0.7×age (2) Electrocardiogram and echocardiogram evaluation Standard 12-lead ECG (Mortara Instrument-Portrait) was performed at supine rest with a paper speed of 25 mm/s and calibration 1 mV/10 mm. The corrected QT interval was calculated using Bazett’s formula. Standard 2D echocardiography (Esaote MyLab 60) was performed by an expert cardiologist. Left ventricular (LV) thickness and diameter were calculated in left parasternal long axis projection. LV ejection fraction and LV end-diastolic volume were calculated as the average of apical four-chamber and apical two-chamber results (biplane evaluation). Endurance training The same CRP was administered to all patients. The CRP consisted of a series of supervised exercise endurance training on cycle ergometer two to three times a week for an overall period between 8 and 12 weeks. In agreement with the international guidelines of exercise prescription, endurance training was established, using C-PET, on the basis of peak VO2. The intensity of endurance training was prescribed, in terms of Watt target, in a range from 40% to 80% of peak VO2, with a duration of 30 minutes per session.17 Training sessions were completed with brief (10 minutes) periods of warm-up and cool-down physical activity (low-intensity aerobic and stretching movements). Responders and non-responders According to a recent meta-analysis on functional capacity response in CAD patients during moderate continuous training, the mean increase of VO2 peak is 2.6 ml/kg/min.19 Therefore, after the C-PET performed at the end of the endurance training, our CAD patients were classified into two groups: patients who responded with an improvement in VO2 greater than the expected value (R group) and those who did not (NR group). Statistics Continuous variables were expressed as mean and standard deviation while categorical variables were expressed as absolute values with percentages. The Student’s t test and Pearson’s χ2 test compared differences between groups for continuous and categorical variables, respectively. A linear mixed-effects model with contrasts a posteriori and unstructured covariance evaluated the difference between R and NR (factor group), the effect of training (factor time) and the interaction between factors. The Lilliefors test was used to test the null hypothesis that data come from a normally distributed population. Residuals analysis was performed to validate the models. For multiple comparisons, the algorithm which controls the expected rate of false-positive results for all positive results (false discovery rate) was used. All analyses were performed with the R: A Language and Environment for Statistical Computing software package (R Core Team, R Foundation for Statistical Computing, Vienna, Austria), setting the statistical significance threshold at P < 0.05. Results Our population of 90 patients (82 men) had an age of 60 ± 11 years. All patients had established ischaemic heart disease (65% had acute coronary syndrome, 35% had stable CAD). Sixty-five patients (72%) were referred to CRP after myocardial revascularisation, obtained by percutaneous transluminal coronary angioplasty in 53% of patients, or by coronary artery bypass graft in 19% of patients. Indices of functional capacity were VO2 peak = 19.0 ± 5.4 ml/kg/min and VO2 predict % = 72.8 ± 18.8%. As described in the methods section, a threshold equal to 2.6 ml/kg/min in the increase of VO2 peak defined R and NR groups. This threshold split our CAD population into 43% of responders and 57% of non-responders. Table 1 compares the two groups at enrollment. The R and NR groups did not differ significantly in terms of anthropometric and clinical variables except for the use of beta-blocker therapy (100% in R vs. 90% in NR). Table 1. Patients’ baseline characteristics: NR and R groups. . R (N = 39) . NR (N = 51) . P value . Age (years) 59 (10) 60 (11) 0.33 Male gender (%) 95% 88% 0.27 Body mass index (kg/m2) 27 (4) 28 (5) 0.12 Family history of CVD (%) 54% 36% 0.09 Hypertension, (%) 59% 74% 0.13 Diabetes type 2 (%) 28% 36% 0.44 Dyslipidaemia (%) 61% 70% 0.40 Smokers (%) 38% 24% 0.14 Leisure-time physical activity 19% 15% 0.56 Laboratory parameters Haemoglobin (g/dL) 13 (2) 14 (2) 0.11 Creatinine (mg/dL) 0.9 (0.2) 1.0 (0.2) 0.25 Drugs Beta-blockers (%) 100% 90% <0.05 Statins (%) 97% 96% 0.72 RAASi (%) 58% 57% 0.84 Indication for cardiac rehabilitation ACS-STEMI (%) 51% 37% 0.18 ACS-NSTEMI (%) 20% 23% 0.73 SCAD (%) 28% 39% 0.28 Type of intervention PTCA (%) 59% 49% 0.35 CABG (%) 23% 16% 0.37 Medical therapy (%) 18% 35% 0.07 ECG PR interval (ms) 169 (21) 167 (27) 0.66 QRS interval (ms) 104 (23) 104 (20) 0.95 Corrected QT interval (ms) 415 (25) 424 (37) 0.21 Cardiac rehabilitation programme Frequency (number of sessions) 26 (9) 24 (7) 0.16 Intensity (W) 82 (30) 80 (24) 0.68 . R (N = 39) . NR (N = 51) . P value . Age (years) 59 (10) 60 (11) 0.33 Male gender (%) 95% 88% 0.27 Body mass index (kg/m2) 27 (4) 28 (5) 0.12 Family history of CVD (%) 54% 36% 0.09 Hypertension, (%) 59% 74% 0.13 Diabetes type 2 (%) 28% 36% 0.44 Dyslipidaemia (%) 61% 70% 0.40 Smokers (%) 38% 24% 0.14 Leisure-time physical activity 19% 15% 0.56 Laboratory parameters Haemoglobin (g/dL) 13 (2) 14 (2) 0.11 Creatinine (mg/dL) 0.9 (0.2) 1.0 (0.2) 0.25 Drugs Beta-blockers (%) 100% 90% <0.05 Statins (%) 97% 96% 0.72 RAASi (%) 58% 57% 0.84 Indication for cardiac rehabilitation ACS-STEMI (%) 51% 37% 0.18 ACS-NSTEMI (%) 20% 23% 0.73 SCAD (%) 28% 39% 0.28 Type of intervention PTCA (%) 59% 49% 0.35 CABG (%) 23% 16% 0.37 Medical therapy (%) 18% 35% 0.07 ECG PR interval (ms) 169 (21) 167 (27) 0.66 QRS interval (ms) 104 (23) 104 (20) 0.95 Corrected QT interval (ms) 415 (25) 424 (37) 0.21 Cardiac rehabilitation programme Frequency (number of sessions) 26 (9) 24 (7) 0.16 Intensity (W) 82 (30) 80 (24) 0.68 Values are expressed as mean and standard deviation or as percentage. CVD: cardiovascular disease; RAASi: renin–angiotensin–aldosterone system inhibitors; ACS-STEMI: acute coronary syndrome-ST elevation myocardial infarction; ACS-NSTEMI: acute coronary syndrome-non-ST elevation myocardial infarction; SCAD: stable coronary artery disease; PTCA: percutaneous transluminal coronary angioplasty; CABG: coronary artery bypass graft; Medical therapy, ischaemic heart disease treated with conservative approach; Leisure-time physical activity: anamnestic relief of at least 150 minutes of physical activity per week before the cardiopulmonary rehabilitation programme. Open in new tab Table 1. Patients’ baseline characteristics: NR and R groups. . R (N = 39) . NR (N = 51) . P value . Age (years) 59 (10) 60 (11) 0.33 Male gender (%) 95% 88% 0.27 Body mass index (kg/m2) 27 (4) 28 (5) 0.12 Family history of CVD (%) 54% 36% 0.09 Hypertension, (%) 59% 74% 0.13 Diabetes type 2 (%) 28% 36% 0.44 Dyslipidaemia (%) 61% 70% 0.40 Smokers (%) 38% 24% 0.14 Leisure-time physical activity 19% 15% 0.56 Laboratory parameters Haemoglobin (g/dL) 13 (2) 14 (2) 0.11 Creatinine (mg/dL) 0.9 (0.2) 1.0 (0.2) 0.25 Drugs Beta-blockers (%) 100% 90% <0.05 Statins (%) 97% 96% 0.72 RAASi (%) 58% 57% 0.84 Indication for cardiac rehabilitation ACS-STEMI (%) 51% 37% 0.18 ACS-NSTEMI (%) 20% 23% 0.73 SCAD (%) 28% 39% 0.28 Type of intervention PTCA (%) 59% 49% 0.35 CABG (%) 23% 16% 0.37 Medical therapy (%) 18% 35% 0.07 ECG PR interval (ms) 169 (21) 167 (27) 0.66 QRS interval (ms) 104 (23) 104 (20) 0.95 Corrected QT interval (ms) 415 (25) 424 (37) 0.21 Cardiac rehabilitation programme Frequency (number of sessions) 26 (9) 24 (7) 0.16 Intensity (W) 82 (30) 80 (24) 0.68 . R (N = 39) . NR (N = 51) . P value . Age (years) 59 (10) 60 (11) 0.33 Male gender (%) 95% 88% 0.27 Body mass index (kg/m2) 27 (4) 28 (5) 0.12 Family history of CVD (%) 54% 36% 0.09 Hypertension, (%) 59% 74% 0.13 Diabetes type 2 (%) 28% 36% 0.44 Dyslipidaemia (%) 61% 70% 0.40 Smokers (%) 38% 24% 0.14 Leisure-time physical activity 19% 15% 0.56 Laboratory parameters Haemoglobin (g/dL) 13 (2) 14 (2) 0.11 Creatinine (mg/dL) 0.9 (0.2) 1.0 (0.2) 0.25 Drugs Beta-blockers (%) 100% 90% <0.05 Statins (%) 97% 96% 0.72 RAASi (%) 58% 57% 0.84 Indication for cardiac rehabilitation ACS-STEMI (%) 51% 37% 0.18 ACS-NSTEMI (%) 20% 23% 0.73 SCAD (%) 28% 39% 0.28 Type of intervention PTCA (%) 59% 49% 0.35 CABG (%) 23% 16% 0.37 Medical therapy (%) 18% 35% 0.07 ECG PR interval (ms) 169 (21) 167 (27) 0.66 QRS interval (ms) 104 (23) 104 (20) 0.95 Corrected QT interval (ms) 415 (25) 424 (37) 0.21 Cardiac rehabilitation programme Frequency (number of sessions) 26 (9) 24 (7) 0.16 Intensity (W) 82 (30) 80 (24) 0.68 Values are expressed as mean and standard deviation or as percentage. CVD: cardiovascular disease; RAASi: renin–angiotensin–aldosterone system inhibitors; ACS-STEMI: acute coronary syndrome-ST elevation myocardial infarction; ACS-NSTEMI: acute coronary syndrome-non-ST elevation myocardial infarction; SCAD: stable coronary artery disease; PTCA: percutaneous transluminal coronary angioplasty; CABG: coronary artery bypass graft; Medical therapy, ischaemic heart disease treated with conservative approach; Leisure-time physical activity: anamnestic relief of at least 150 minutes of physical activity per week before the cardiopulmonary rehabilitation programme. Open in new tab Moreover, R and NR patients followed training programmes with similar frequency and intensity, i.e. on average 25 ± 8 sessions at a load of 80 ± 27 Watts. Figure 1 shows the distributions of increments in VO2 peak separately in the two groups: increment distributions are centred on zero in the NR group. Figure 1. Open in new tabDownload slide Distributions of increments (Δ = after–before endurance training) in functional capacity over our coronary artery disease (CAD) patients. Left panel: oxygen uptake (VO2) peak. Right panel: VO2 peak (%), i.e. VO2 peak as a percentage of the theoretical value for an individual with the same age, gender, height and weight of the patient. Dashed/open boxes indicate patients classified as NR or R, respectively, according to ΔVO2 peak greater or lower than 2.6 ml/kg/min. Distributions of R and NR patients may overlap when functional capacity is expressed as ΔVO2 peak (%); this happens in very few cases only, suggesting the equivalence of VO2 peak and VO2 peak (%) indices for identifying R and NR subgroups. The autonomic indices of fitness derived from C-PET are shown in Figure 2. For both HRR and CR, the interaction between factors was significant (P < 0.05 for HR, P < 0.01 for CR), indicating that exercise training had different effects on R and NR patients. Post-hoc analysis provided evidence that training increased significantly both the autonomic indices of fitness in the R group (HRR: from 16.9 ± 7.0 to 21.0 ± 8.7 bpm; CR: from 45.1 ± 16.9% to 72.7 ± 34.1%) while HRR and CR remained unchanged in the NR group (HRR: from 15.2 ± 9.9 to 15.8 ± 8.5 bpm; CR: from 49.3 ± 18.6% to 48.2 ± 36.5%). Consequently, while before training HRR and CR were similar in the two groups, after training HRR and CR were significantly higher in R than in NR patients. Percentage variations of HRR and CR as a continuous function of VO2 peak increments are provided by Supplementary Figure 1. Figure 2. Open in new tabDownload slide Autonomic indices of fitness before and after training (mean and its 95% confidence interval); ** indicates differences after training significant at P < 0.01; # and ## indicate differences between R and NR significant at P < 0.05 and P < 0.01. Figure 3 illustrates the effects of CRP on echocardiographic variables. Exercise training tended to increase the LV ejection fraction in all patients (from 46.6 ± 13.0% to 48.1 ± 11.3%, significance of factor time P = 0.05), without evident differences between groups (significance of factors interaction P = 0.14). By contrast, exercise training produced very different effects on the LV end-diastolic volume (significance of factors interaction P < 0.001), which decreased markedly in R (from 157.2 ± 61.4 to 142.3 ± 48.3 mL, P < 0.01), without significant changes in NR (143.3 ± 55.4 vs. 157.9 ± 55.4 mL, P = 0.12). Table 2 reports the effects of training on cardiovascular and ventilatory variables during C-PET. No factors were statistically significant for any variable but the interaction between factors for resting HR (P < 0.01) and resting HR increased significantly after training in NR patients only. Table 2. C-PET variables in R and NR groups before and after training. . Before training . After training . . R . NR . R . NR . SAP rest (mmHg) 115.9 (13.2) 121.3 (15.0) 117.5 (12.3) 117.1 (14.1) SAP peak (mmHg) 170.6 (33.4) 174.5 (28.7) 178.3 (30) 168.3 (31.2) DAP rest (mmHg) 68.6 (8.1) 69.6 (8.5) 69.3 (9.2) 67.4 (7.8) DAP peak (mmHg) 87.6 (14.2) 86.3 (11.1) 86.3 (12.6) 85.8 (12.7) HR rest (bpm) 67.2 (13.6) 60.1 (13.1)§ 63.7 (10.8) 66.6 (11.4)** HR peak (bpm) 112.7 (19.0) 112.5 (21.0) 117.3 (19.3) 112.2 (20.0) Peak VO2 (ml/kg/min) 18.7 (5.7) 19.2 (5.2) 24.1 (6.1)** 19.1 (4.7)§§ Peak VO2 (%) 69.1 (19) 75.7 (18.3) 89.2 (20.6)** 75.6 (16.1)§§ RER peak 1.03 (0.10) 0.99 (0.11) 1.03 (0.12) 1.02 (0.09) VE/VO2 peak 37.5 (7.1) 37.8 (8.9) 37.3 (10.2) 38.8 (8.9) VE/VCO2 peak 36.4 (6.2) 38.2 (8.0) 36.2 (8.8) 38.0 (8.0) VE/VCO2 slope 34.1 (9.0) 35.3 (8.9) 33 (9.4) 35.4 (8.9) VO2/W slope 9.9 (4.5) 9.3 (2.0) 9.5 (1.8) 8.8 (1.7) . Before training . After training . . R . NR . R . NR . SAP rest (mmHg) 115.9 (13.2) 121.3 (15.0) 117.5 (12.3) 117.1 (14.1) SAP peak (mmHg) 170.6 (33.4) 174.5 (28.7) 178.3 (30) 168.3 (31.2) DAP rest (mmHg) 68.6 (8.1) 69.6 (8.5) 69.3 (9.2) 67.4 (7.8) DAP peak (mmHg) 87.6 (14.2) 86.3 (11.1) 86.3 (12.6) 85.8 (12.7) HR rest (bpm) 67.2 (13.6) 60.1 (13.1)§ 63.7 (10.8) 66.6 (11.4)** HR peak (bpm) 112.7 (19.0) 112.5 (21.0) 117.3 (19.3) 112.2 (20.0) Peak VO2 (ml/kg/min) 18.7 (5.7) 19.2 (5.2) 24.1 (6.1)** 19.1 (4.7)§§ Peak VO2 (%) 69.1 (19) 75.7 (18.3) 89.2 (20.6)** 75.6 (16.1)§§ RER peak 1.03 (0.10) 0.99 (0.11) 1.03 (0.12) 1.02 (0.09) VE/VO2 peak 37.5 (7.1) 37.8 (8.9) 37.3 (10.2) 38.8 (8.9) VE/VCO2 peak 36.4 (6.2) 38.2 (8.0) 36.2 (8.8) 38.0 (8.0) VE/VCO2 slope 34.1 (9.0) 35.3 (8.9) 33 (9.4) 35.4 (8.9) VO2/W slope 9.9 (4.5) 9.3 (2.0) 9.5 (1.8) 8.8 (1.7) Values are expressed as mean and standard deviation. ** indicates significance P < 0.01 for after training vs. before training comparison. § and §§ indicate significances P < 0.05 and P < 0.01, respectively, for R vs. NR comparison. SAP: systolic arterial pressure; DAP: diastolic arterial pressure; RER: respiratory exchange ratio. See text for abbreviations. Open in new tab Table 2. C-PET variables in R and NR groups before and after training. . Before training . After training . . R . NR . R . NR . SAP rest (mmHg) 115.9 (13.2) 121.3 (15.0) 117.5 (12.3) 117.1 (14.1) SAP peak (mmHg) 170.6 (33.4) 174.5 (28.7) 178.3 (30) 168.3 (31.2) DAP rest (mmHg) 68.6 (8.1) 69.6 (8.5) 69.3 (9.2) 67.4 (7.8) DAP peak (mmHg) 87.6 (14.2) 86.3 (11.1) 86.3 (12.6) 85.8 (12.7) HR rest (bpm) 67.2 (13.6) 60.1 (13.1)§ 63.7 (10.8) 66.6 (11.4)** HR peak (bpm) 112.7 (19.0) 112.5 (21.0) 117.3 (19.3) 112.2 (20.0) Peak VO2 (ml/kg/min) 18.7 (5.7) 19.2 (5.2) 24.1 (6.1)** 19.1 (4.7)§§ Peak VO2 (%) 69.1 (19) 75.7 (18.3) 89.2 (20.6)** 75.6 (16.1)§§ RER peak 1.03 (0.10) 0.99 (0.11) 1.03 (0.12) 1.02 (0.09) VE/VO2 peak 37.5 (7.1) 37.8 (8.9) 37.3 (10.2) 38.8 (8.9) VE/VCO2 peak 36.4 (6.2) 38.2 (8.0) 36.2 (8.8) 38.0 (8.0) VE/VCO2 slope 34.1 (9.0) 35.3 (8.9) 33 (9.4) 35.4 (8.9) VO2/W slope 9.9 (4.5) 9.3 (2.0) 9.5 (1.8) 8.8 (1.7) . Before training . After training . . R . NR . R . NR . SAP rest (mmHg) 115.9 (13.2) 121.3 (15.0) 117.5 (12.3) 117.1 (14.1) SAP peak (mmHg) 170.6 (33.4) 174.5 (28.7) 178.3 (30) 168.3 (31.2) DAP rest (mmHg) 68.6 (8.1) 69.6 (8.5) 69.3 (9.2) 67.4 (7.8) DAP peak (mmHg) 87.6 (14.2) 86.3 (11.1) 86.3 (12.6) 85.8 (12.7) HR rest (bpm) 67.2 (13.6) 60.1 (13.1)§ 63.7 (10.8) 66.6 (11.4)** HR peak (bpm) 112.7 (19.0) 112.5 (21.0) 117.3 (19.3) 112.2 (20.0) Peak VO2 (ml/kg/min) 18.7 (5.7) 19.2 (5.2) 24.1 (6.1)** 19.1 (4.7)§§ Peak VO2 (%) 69.1 (19) 75.7 (18.3) 89.2 (20.6)** 75.6 (16.1)§§ RER peak 1.03 (0.10) 0.99 (0.11) 1.03 (0.12) 1.02 (0.09) VE/VO2 peak 37.5 (7.1) 37.8 (8.9) 37.3 (10.2) 38.8 (8.9) VE/VCO2 peak 36.4 (6.2) 38.2 (8.0) 36.2 (8.8) 38.0 (8.0) VE/VCO2 slope 34.1 (9.0) 35.3 (8.9) 33 (9.4) 35.4 (8.9) VO2/W slope 9.9 (4.5) 9.3 (2.0) 9.5 (1.8) 8.8 (1.7) Values are expressed as mean and standard deviation. ** indicates significance P < 0.01 for after training vs. before training comparison. § and §§ indicate significances P < 0.05 and P < 0.01, respectively, for R vs. NR comparison. SAP: systolic arterial pressure; DAP: diastolic arterial pressure; RER: respiratory exchange ratio. See text for abbreviations. Open in new tab The effects of training on autonomic and echocardiographic indices did not change, removing the five patients in our study not treated with beta-blockers (see Supplementary Table 1). Discussion Our study represents a demonstration that moderate-intensity continuous training improves aerobic functional capacity and that this improvement is strikingly associated with positive adaptations (namely increased cardiac vagal modulation) of autonomic function in CAD patients with substantially preserved LV ejection fraction. In addition, the present work provides further evidence of a beneficial effect of exercise training on autonomic tone in CAD patients, this improvement being obtained by means of two indices easily measured with a symptom-limited exercise testing: CR and HRR. As changes in aerobic capacity and autonomic function after exercise training have generally been separately examined in patients with cardiovascular disease,20,21 our study provides novel evidence on a potential role of cardiac autonomic adaptations in determining different responses in functional capacity in CAD patients after the same rehabilitation programme (continuous moderate training). While the beneficial effect of regular aerobic exercise on functional capacity and sympatho-vagal balance responses has been particularly investigated in heart failure patients with either preserved22 or reduced7,23,24 LV ejection fraction, only few data have been obtained on a possible relationship between aerobic fitness and autonomic function in CAD patients undergoing moderate continuous training. In fact, our results clearly demonstrate that CRP does not invariably improve functional capacity and autonomic markers such as CR and HRR in all patients with established CAD. In this regard, our data may also help to interpret the results of a recent study that reported only a small non-significant increase in the vagal modulation of heart rate, as assessed by heart rate variability analysis during supine rest, in CAD patients after exercise-based cardiac rehabilitation.25 Actually, our study indicates that the increased cardiac vagal modulation occurs in the R group only, which represents less than half the population of our CAD patients. A systematic review (eight studies including 449 patients) evaluated the relationship between aerobic training and HRR, showing level 1 A evidence that aerobic exercise training increases HRR in patients with established ischaemic heart disease,6 and an improvement of HRR and CR has also been described in patients with heart failure undergoing exercise training.26 More recently, a meta-analysis summarising the effects of exercise training on functional capacity responses in CAD patients19 has shown a mean increase of VO2 peak of 2.6 ml/kg/min during moderate continuous training in these subjects. Accordingly, this value set the threshold for dividing our CAD patients into the R and NR groups. This allowed us to demonstrate that only patients with VO2 peak increase greater than 2.6 ml/kg/min (R group) exhibited a significant improvement of their autonomic indices of aerobic fitness. Unlike the above-mentioned studies investigating VO2 peak and autonomic function separately, our results have been obtained through an integrated analysis of both functional capacity and autonomic markers during a CRP with similar frequency and intensity. The same integrated approach, performed in 30 patients with CHF,7 showed that only a 3-month supervised continuous training programme rather than interval training improved both exercise capacity and autonomic function, expressed as CR and HRR. Moreover, in 57 patients with advanced CHF it has been demonstrated that a major degree of autonomic dysfunction, evaluated by heart rate variability, did result in a lesser improvement of functional capacity after a short period (2 weeks) of exercise-based rehabilitation.23 Furthermore, it has previously been shown that in heart failure patients with severely reduced systolic function a reversed LV remodelling occurred after aerobic interval training, and not after moderate continuous training.27 At variance with the above-mentioned results, our demonstration of a reduced LV end-diastolic volume in the R group suggests that moderate continuous training may also play a role in determining a reversed cardiac remodelling in established CAD patients with substantially preserved ejection fraction. On the other hand, in a recent trial including 200 CAD patients randomly assigned to a supervised 12-week CRP, a similar improvement in exercise capacity with both aerobic continuous and interval training has been observed.28 Figure 3. Open in new tabDownload slide Echocardiographic indices before and after training (mean and its 95% confidence interval); LVEF: left ventricular ejection fraction; LVEDV: left ventricular end-diastolic volume; IVSd: interventricular septal thickness at end-diastole; ** indicates a difference after training significant at P < 0.01. In conclusion, our results suggest a possible contributing role of autonomic function in aerobic capacity improvement in patients with established CAD undergoing exercise-based cardiac rehabilitation. Given that both aerobic capacity and autonomic function alterations are powerful predictors of cardiac and overall mortality, an integrated evaluation of these parameters may help physicians to identify patients with higher cardiovascular risk and, in the same way, to improve both exercise prescription and pharmacological approach. Limitations Because of the retrospective nature of this analysis, the present study should be considered hypothesis generating. It was designed to detect an association and not a causal relationship between aerobic capacity and CR or HRR improvements during moderate continuous training. It should also be considered that measures other than CR and HRR might have adequately described the autonomic function,29 by differently showing evidence of the changes in the autonomic cardiac control induced by endurance training. However, as CR and HRR directly reflect the autonomic response to an exercise load, these indices would appear particularly appropriate for describing the capability of cardio-rehabilitation exercise-based programmes to improve autonomic cardiovascular control. Author contribution DL, MB, UC and PCo contributed to the conception, design, acquisition, analysis and interpretation of data for the work. PCa contributed to the analysis and interpretation of data for the work. CSC contributed to the acquisition of data for the work. AF contributed to statistical analysis of data. MFP, PTU, VB and LB contributed to the interpretation of data for the work. DL, MB, UC and PCo drafted the manuscript and PCa contributed to critically revising it. All authors gave final approval and agree to be accountable for all aspects of work ensuring integrity and accuracy. 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 Giannuzzi P , Saner H, Bjornstad Het al. Working Group on Cardiac Rehabilitation, Exercise Physiology of the European Society of Cardiology . Secondary prevention through cardiac rehabilitation: position paper of the Working Group on Cardiac Rehabilitation and Exercise Physiology of the European Society of Cardiology . Eur Heart J 2003 ; 24 : 1273 – 1278 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Graham I , Atar D, Borch-Johnsen Ket al. ESC Committee for Practice Guidelines . European guidelines on cardiovascular disease prevention in clinical practice: executive summary . Atherosclerosis 2007 ; 194 : 1 – 45 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Iliou MC , Pavy B, Martinez Jet al. Exercise training is safe after coronary stenting: a prospective multicentre study . Eur J Prev Cardiol 2015 ; 22 : 27 – 34 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Belardinelli R , Georgiou D, Cianci Get al. Randomized, controlled trial of long-term moderate exercise training in chronic heart failure. Effects of functional capacity, quality of life, and clinical outcome . Circulation 1999 ; 99 : 1173 – 1182 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Coats AJS , Adamopoulos S, Radaelli Aet al. Controlled trial of physical training in chronic heart failure. Exercise performance, hemodynamics, ventilation, and autonomic function . Circulation 1992 ; 85 : 2119 – 2131 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Snoek JA , van Berkel S, van Meeteren Net al. Effect of aerobic training on heart rate recovery in patients with established heart disease; a systematic review . Plos One 2013 ; 8 : e83907 – e83907 . DOI: 10.1371 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Dimopoulos S , Anastasiou-Nana M, Sakellariou Det al. Effects of exercise rehabilitation program on heart rate recovery in patients with chronic heart failure . Eur J Cardiovasc Prev Rehab 2006 ; 13 : 67 – 73 . Google Scholar Crossref Search ADS WorldCat 8 Brubaker PH , Kitzman DW. Chronotropic incompetence: causes, consequences, and management . Circulation 2011 ; 123 : 1010 – 1020 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Arai Y , Saul JP, Albrecht Pet al. Modulation of cardiac autonomic activity during and immediately after exercise . Am J Physiol 1989 ; 256 : H132 – H141 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 10 Imai K , Sato H, Hori Met al. Vagally mediated heart rate recovery after exercise is accelerated in athletes but blunted in patients with chronic heart failure . J Am Coll Cardiol 1994 ; 24 : 1529 – 1535 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Colucci WS , Ribeiro JP, Rocco MBet al. Impaired chronotropic response to exercise in patients with congestive heart failure. Role of postsynaptic β-adrenergic desensitization . Circulation 1989 ; 80 : 314 – 323 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Shetler K , Marcus R, Froelicher VFet al. Heart rate recovery: validation and methodologic issues . J Am Coll Cardiol 2001 ; 38 : 1980 – 1987 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Chen MS , Blackstone EH, Pothier CEet al. Heart rate recovery and impact of myocardial revascularization on long-term mortality . Circulation 2004 ; 110 : 2851 – 2857 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Hautala AJ , Makikallio TH, Kiviniemi Aet al. Cardiovascular autonomic function correlates with the response to aerobic training in healthy sedentary subjects . Am J Physiol Heart Circ Physiol 2003 ; 285 : H1747 – H1752 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Hautala AJ , Kiviniemi AM, Tulppo MP. Individual responses to aerobic exercise: the role of the autonomic nervous system . Neurosci Biobehav Rev 2009 ; 33 : 107 – 115 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Borg GA . Psychophysical bases of perceived exertion . Med Sci Sports Exerc 1982 ; 14 : 377 – 381 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 17 Fetcher GF , Ades PA, Kligfield P. Exercise standards for testing and training: a scientific statement from the American Heart Association . Circulation 2013 ; 128 : 873 – 934 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Cole CR , Blackstone EH, Pashkow FJet al. Heart-rate recovery immediately after exercise as a predictor of mortality . N Engl J Med 1999 ; 341 : 1351 – 1357 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Pattyn N , Coeckelberghs E, Buys Ret al. Aerobic interval training vs moderate continuous training in coronary artery disease patients: a systematic review and meta-analysis . Sports Med 2014 ; 44 : 687 – 700 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Hautala AJ , Kiviniemi AM, Mäkikallio THet al. Individual differences in the responses to endurance and resistance training . Eur J Appl Physiol 2006 ; 96 : 535 – 542 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Legramante JM , Iellamo F, Massaro Met al. Effects of residential exercise training on heart rate recovery in coronary artery patients . Am J Physiol Heart Circ Physiol 2007 ; 292 : H510 – H515 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Pandey A , Parashar A, Kumbhani DJet al. Exercise training in patients with heart failure and preserved ejection fraction. Meta-analysis of randomized control trials . Circ Heart Fail 2015 ; 8 : 33 – 40 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Compostella L , Russo N, Setzu Tet al. Autonomic dysfunction predicts poor physical improvement after cardiac rehabilitation in patients with heart failure . Res Cardiovasc Med 2014 ; 3 ( 4 ): e25237 – e25237 . DOI: 10.5812/cardiovascmed.25237 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Spee RF , Niemeijer VM, Wijn PFet al. Effects of high-intensity interval training on central haemodynamics and skeletal muscle oxygenation during exercise in patients with chronic heart failure . Eur J Prev Cardiol 2016 ; 23 : 1943 – 1952 . . DOI: 10.1177/2047487316661615 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Oliveira NL , Ribeiro F, Teixeira Met al. Effect of 8-week exercise-based cardiac rehabilitation on cardiac autonomic function: a randomized controlled trial in myocardial infarction patients . Am Heart J 2014 ; 167 : 753–761.e3. Google Scholar OpenURL Placeholder Text WorldCat 26 Keteyian SJ , Brawner CA, Schairer JRet al. Effects of exercise training on chronotropic incompetence in patients with heart failure . Am Heart J 1999 ; 138 : 233 – 240 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Wisløff U , Støylen A, Loennechen JPet al. Superior cardiovascular effect of aerobic interval training versus moderate continuous training in heart failure patients. A randomized study . Circulation 2007 ; 115 : 3086 – 3094 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Conraads VM , Pattyn N, De Maeyer C. Aerobic interval training and continuous training equally improve aerobic exercise capacity in patients with coronary artery disease: the SAINTEX-CAD study . Int J Cardiol 2015 ; 179 : 203 – 210 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology . Heart rate variability: standards of measurement, physiological interpretation and clinical use . Circulation 1996 ; 93 : 1043 – 1065 . Crossref Search ADS PubMed WorldCat © The European Society of Cardiology 2017 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 2017

Abstract Background Exercise training improves neurovascular control and functional capacity in heart failure (HF) patients. However, the influence of the aetiology on these benefits is unknown. We compared the effects of exercise training on neurovascular control and functional capacity in idiopathic, ischaemic and hypertensive HF patients. Design Subjects consisted of 45 exercise-trained HF patients from our database (2000–2015), aged 40–70 years old, functional class II/III and ejection fraction ≤40%, and they were divided into three groups: idiopathic (n = 11), ischaemic (n = 18) and hypertensive (n = 16). Methods Functional capacity was determined by cardiopulmonary exercise testing. Muscle sympathetic nerve activity (MSNA) was recorded by microneurography. Forearm blood flow (FBF) was measured by venous occlusion plethysmography. Results Four months of exercise training significantly reduced MSNA and significantly increased FBF in all groups. However, the relative reduction in MSNA was greater in hypertensive patients compared with that in idiopathic patients (frequency: −34% vs. −15%, p = 0.01; incidence: −31% vs. −12%, p = 0.02). No differences were found between hypertensive patients and ischaemic patients. The relative increase in FBF was greater in hypertensive patients than in ischaemic and idiopathic patients (42% vs. 15% and 17%, respectively, p = 0.02). The relative increase in forearm vascular conductance was greater in hypertensive patients compared with those in ischaemic and idiopathic patients (57% vs. 13% and 26%, respectively, p = 0.001). Exercise training significantly and similarly increased peak oxygen consumption in all groups. Conclusion The exercise-induced improvement in neurovascular control is more pronounced in hypertensive HF patients than in idiopathic and ischaemic HF patients. The increase in functional capacity is independent of aetiology. Heart failure, exercise training, forearm blood flow, muscle sympathetic nerve activity Introduction Despite advances in pharmacological treatment, the prevalence of heart failure (HF) continues to rise worldwide.1 Accumulated evidence shows that HF is characterised by intense peripheral vasoconstriction associated with neurohumoral abnormalities.2 In the chronic stage, these alterations contribute to skeletal myopathy, which explains, in great part, the exercise intolerance in HF patients.2 Myocardial infarction, hypertension, and aetiology are the major causes of HF.1,3,4 In Western societies, ischaemic disease remains the most prevalent cause of HF. In addition, patients suffering from ischaemic disease have lower survival rates and exercise capacity than non-ischaemic patients.5 These findings provide evidence that the aetiology can influence the prognosis of HF patients. The American Heart Association and American College of Cardiology guidelines recommend exercise training as an adjuvant therapy in the treatment of HF.6 Exercise training remarkably improves neurovascular control in HF patients.2 We have consistently reported that supervised exercise training reduces muscle sympathetic nerve activity (MSNA) and increases forearm blood flow (FBF) in HF.7–13 These benefits provoked by exercise training occur regardless of sex and age.11,12 The reduction in vascular constriction contributes to the improvement in skeletal muscle myopathy, which in turn plays an important role in the amelioration of functional capacity and quality of life in HF patients.2 In addition, recent reports demonstrate that exercise training decreases hospital admissions for the treatment of HF.14–16 However, the influence of the aetiology on the neurovascular control and peripheral vasoconstriction in exercise-trained chronic HF patients remains unknown. In the present study, we investigated the effects of exercise training on MSNA, FBF and functional capacity in ischaemic, hypertensive and idiopathic HF patients. Methods Study population Patients with clinically stable idiopathic, ischaemic and hypertensive HF, aged 40–70 years, functional class II and III and left ventricular ejection fraction (LVEF) ≤40% were included in the study. Patients with unstable angina, a recent myocardial infarction (<3 months), chronic obstructive pulmonary disease, on dialysis or with neurological and/or orthopaedic disabilities were excluded from the study. Study protocol Forty-five consecutive outpatients with HF from the database of the Unit of Cardiovascular Rehabilitation and Physiology Exercise, Heart Institute, University of São Paulo Medical School, who had completed the exercise sessions from 2000 to 2015, were divided into three groups: ischaemic (n = 18), hypertensive (n = 16) and idiopathic (n = 11). All patients were evaluated at baseline and after 4 months of supervised exercise training. The Human Subject Protection Committee of the Heart Institute (SDC-4352/16/018) and Clinical Hospital, University of São Paulo Medical School (CAPPesq-5946516.3.0000.0068) approved the study. Methods and procedures Cardiopulmonary exercise testing Maximal exercise capacity was determined by a maximal progressive exercise test on a cycle ergometer using a ramp protocol of 5–10 W.8 MSNA evaluation MSNA was recorded directly from the peroneal nerve using a microneurography technique.8 MSNA bursts were expressed as burst frequency (bursts/minute) and burst incidence (bursts/100 heartbeats (HB)). FBF evaluation FBF was measured using venous occlusion plethysmography.8 Forearm vascular conductance (FVC) was calculated by dividing the FBF by the mean arterial pressure, multiplied by 100. FBF was expressed as mL/minute/100 mL tissue and FVC was expressed in arbitrary units. Other evaluations Arterial pressure was monitored non-invasively and intermittently from an automatic and oscillometric cuff. Heart rate (HR) was monitored continuously through an electrocardiogram. LVEF was determined by 2D echocardiography using Simpson’s method. Exercise training The exercise training protocol was based on supervised moderate aerobic exercise training at the Heart Institute. The 4–month training programme consisted of three 60–minute exercise sessions/week. Each session consisted of 5 minutes of stretching exercises, 40 minutes of cycling, 10 minutes of local strengthening exercises and 5 minutes of cooldown. The exercise intensity was established by HR levels that corresponded to anaerobic thresholds up to 10% below the respiratory compensation point obtained in the cardiopulmonary exercise testing. Aerobic exercise duration increased progressively so that all patients could perform 40 minutes of cycling at the established intensity.13 Statistical analysis The data are presented as means ± SE. The Kolmogorov–Smirnov and Levene’s tests were used to assess normality of distribution and homogeneity. The baseline characteristics were tested by one-way analysis of variance (ANOVA). Paired Student’s t-tests were used to assess the within-group differences. The possible differences in exercise training responses (Δ%) among groups were tested by one-way ANOVA. In the case of significance, Scheffé’s post hoc multiple comparisons were conducted. Non-parametric tests were used when needed (Kruskal–Wallis test, Mann–Whitney U–test and Wilcoxon matched pairs test). A χ2 test was used to assess categorical data differences. Pearson’s coefficient was used to assess the relationships between percentage changes in MSNA, FBF, FVC and peak VO2. Probability values of p < 0.05 were considered statistically significant. Results Baseline characteristics The baseline characteristics are displayed in Table 1. There were no differences between groups in terms of age, body mass index, functional class, β-blocker use, angiotensin-converting enzyme inhibitor/angiotensin II receptor antagonist use, spironolactone use, diabetes, dyslipidaemia, peak VO2, LVEF, HR, MSNA, FBF and FVC. There were more men in the ischaemic group than in the idiopathic and hypertensive groups. Hypertensive patients used more diuretics than idiopathic and ischaemic patients (100% vs. 64% and 83%, respectively), whereas ischaemic patients used more statins than idiopathic and hypertensive patients (78% vs. 45% and 31%, respectively). There were more hypertensive patients in the ischaemic and hypertensive groups than in the idiopathic group (61% and 100%, respectively, vs. 36%). Systolic, diastolic and mean arterial blood pressures were significantly higher in the hypertensive group compared with that in the idiopathic and ischaemic groups. Table 1. Baseline characteristics in patients with heart failure. . Idiopathic (n = 11) . Ischaemic (n = 18) . Hypertensive (n = 16) . p–value . Age (years) 56 ± 3 59 ± 1 62 ± 2 0.16 BMI (kg/m2) 27 ± 1 28 ± 1 27 ± 1 0.66 Sex  Male 5 18 11 0.03  Female 6 0 5 Functional class  NYHA-II 9 7 9 0.08  NYHA-III 2 11 7 Medications  β-blocker 11 18 16 1.00  ACEI/ARA 11 18 16 1.00  Spironolactone 7 15 13 0.43  Diuretics 7 15 16 0.04  Statins 5 14 5 0.02 Diabetes 1 5 2 0.35 Hypertension 4 11 16 0.001 Dyslipidaemia 7 14 9 0.40 Peak VO2 (mL/kg/minute) 16 ± 1 16 ± 1 15 ± 1 0.75 LVEF (%) 27 ± 2 27 ± 2 31 ± 2 0.12 Heart rate (bpm) 67 ± 3 64 ± 2 65 ± 2 0.70 SBP (mmHg) 122 ± 5 113 ± 4 138 ± 5a 0.004 DBP (mmHg) 68 ± 2 63 ± 2 77 ± 4a 0.02 MBP (mmHg) 88 ± 4 82 ± 2 95 ± 4a 0.03 MSNA (burst/minute) 39 ± 3 48 ± 3 45 ± 3 0.15 MSNA (burst/100 HB) 59 ± 5 75 ± 4 70 ± 5 0.07 FBF (mL/minute/100 mL) 1.62 ± 0.09 1.63 ± 0.08 1.56 ± 0.11 0.85 FVC (a.u.) 1.85 ± 0.10 2.03 ± 0.12 1.67 ± 0.11 0.09 . Idiopathic (n = 11) . Ischaemic (n = 18) . Hypertensive (n = 16) . p–value . Age (years) 56 ± 3 59 ± 1 62 ± 2 0.16 BMI (kg/m2) 27 ± 1 28 ± 1 27 ± 1 0.66 Sex  Male 5 18 11 0.03  Female 6 0 5 Functional class  NYHA-II 9 7 9 0.08  NYHA-III 2 11 7 Medications  β-blocker 11 18 16 1.00  ACEI/ARA 11 18 16 1.00  Spironolactone 7 15 13 0.43  Diuretics 7 15 16 0.04  Statins 5 14 5 0.02 Diabetes 1 5 2 0.35 Hypertension 4 11 16 0.001 Dyslipidaemia 7 14 9 0.40 Peak VO2 (mL/kg/minute) 16 ± 1 16 ± 1 15 ± 1 0.75 LVEF (%) 27 ± 2 27 ± 2 31 ± 2 0.12 Heart rate (bpm) 67 ± 3 64 ± 2 65 ± 2 0.70 SBP (mmHg) 122 ± 5 113 ± 4 138 ± 5a 0.004 DBP (mmHg) 68 ± 2 63 ± 2 77 ± 4a 0.02 MBP (mmHg) 88 ± 4 82 ± 2 95 ± 4a 0.03 MSNA (burst/minute) 39 ± 3 48 ± 3 45 ± 3 0.15 MSNA (burst/100 HB) 59 ± 5 75 ± 4 70 ± 5 0.07 FBF (mL/minute/100 mL) 1.62 ± 0.09 1.63 ± 0.08 1.56 ± 0.11 0.85 FVC (a.u.) 1.85 ± 0.10 2.03 ± 0.12 1.67 ± 0.11 0.09 Values are mean ± SE. a Versus ischaemic heart failure patients. BMI: body mass index; NYHA: New York Heart Association; ACEI: angiotensin-converting enzyme inhibitor; ARA: angiotensin II receptor antagonist; VO2: oxygen uptake; LVEF: left ventricular ejection fraction; SBP: systolic blood pressure; DBP: diastolic blood pressure; MBP: mean blood pressure; MSNA: muscle sympathetic nerve activity; HB: heartbeats; FBF: forearm blood flow; FVC: forearm vascular conductance. Open in new tab Table 1. Baseline characteristics in patients with heart failure. . Idiopathic (n = 11) . Ischaemic (n = 18) . Hypertensive (n = 16) . p–value . Age (years) 56 ± 3 59 ± 1 62 ± 2 0.16 BMI (kg/m2) 27 ± 1 28 ± 1 27 ± 1 0.66 Sex  Male 5 18 11 0.03  Female 6 0 5 Functional class  NYHA-II 9 7 9 0.08  NYHA-III 2 11 7 Medications  β-blocker 11 18 16 1.00  ACEI/ARA 11 18 16 1.00  Spironolactone 7 15 13 0.43  Diuretics 7 15 16 0.04  Statins 5 14 5 0.02 Diabetes 1 5 2 0.35 Hypertension 4 11 16 0.001 Dyslipidaemia 7 14 9 0.40 Peak VO2 (mL/kg/minute) 16 ± 1 16 ± 1 15 ± 1 0.75 LVEF (%) 27 ± 2 27 ± 2 31 ± 2 0.12 Heart rate (bpm) 67 ± 3 64 ± 2 65 ± 2 0.70 SBP (mmHg) 122 ± 5 113 ± 4 138 ± 5a 0.004 DBP (mmHg) 68 ± 2 63 ± 2 77 ± 4a 0.02 MBP (mmHg) 88 ± 4 82 ± 2 95 ± 4a 0.03 MSNA (burst/minute) 39 ± 3 48 ± 3 45 ± 3 0.15 MSNA (burst/100 HB) 59 ± 5 75 ± 4 70 ± 5 0.07 FBF (mL/minute/100 mL) 1.62 ± 0.09 1.63 ± 0.08 1.56 ± 0.11 0.85 FVC (a.u.) 1.85 ± 0.10 2.03 ± 0.12 1.67 ± 0.11 0.09 . Idiopathic (n = 11) . Ischaemic (n = 18) . Hypertensive (n = 16) . p–value . Age (years) 56 ± 3 59 ± 1 62 ± 2 0.16 BMI (kg/m2) 27 ± 1 28 ± 1 27 ± 1 0.66 Sex  Male 5 18 11 0.03  Female 6 0 5 Functional class  NYHA-II 9 7 9 0.08  NYHA-III 2 11 7 Medications  β-blocker 11 18 16 1.00  ACEI/ARA 11 18 16 1.00  Spironolactone 7 15 13 0.43  Diuretics 7 15 16 0.04  Statins 5 14 5 0.02 Diabetes 1 5 2 0.35 Hypertension 4 11 16 0.001 Dyslipidaemia 7 14 9 0.40 Peak VO2 (mL/kg/minute) 16 ± 1 16 ± 1 15 ± 1 0.75 LVEF (%) 27 ± 2 27 ± 2 31 ± 2 0.12 Heart rate (bpm) 67 ± 3 64 ± 2 65 ± 2 0.70 SBP (mmHg) 122 ± 5 113 ± 4 138 ± 5a 0.004 DBP (mmHg) 68 ± 2 63 ± 2 77 ± 4a 0.02 MBP (mmHg) 88 ± 4 82 ± 2 95 ± 4a 0.03 MSNA (burst/minute) 39 ± 3 48 ± 3 45 ± 3 0.15 MSNA (burst/100 HB) 59 ± 5 75 ± 4 70 ± 5 0.07 FBF (mL/minute/100 mL) 1.62 ± 0.09 1.63 ± 0.08 1.56 ± 0.11 0.85 FVC (a.u.) 1.85 ± 0.10 2.03 ± 0.12 1.67 ± 0.11 0.09 Values are mean ± SE. a Versus ischaemic heart failure patients. BMI: body mass index; NYHA: New York Heart Association; ACEI: angiotensin-converting enzyme inhibitor; ARA: angiotensin II receptor antagonist; VO2: oxygen uptake; LVEF: left ventricular ejection fraction; SBP: systolic blood pressure; DBP: diastolic blood pressure; MBP: mean blood pressure; MSNA: muscle sympathetic nerve activity; HB: heartbeats; FBF: forearm blood flow; FVC: forearm vascular conductance. Open in new tab Effects of exercise training Exercise training significantly and similarly increased peak VO2 in idiopathic (p < 0.001), ischaemic (p < 0.01) and hypertensive (p < 0.001) groups (Tables 2 and 3). Exercise training caused no significant changes in resting HR in all three groups (Tables 2 and 3). Systolic arterial pressure levels were higher in exercise-trained ischaemic patients (p = 0.02). No significant changes were observed in idiopathic and hypertensive patients (Table 2 and 3). Exercise training significantly reduced diastolic and mean arterial pressure levels in hypertensive patients (p = 0.002 and p = 0.02, respectively). No changes were found in ischaemic and idiopathic patients (Tables 2 and 3). Table 2. Effects of exercise training on physiologic parameters. . Idiopathic (n = 11) . Ischaemic (n = 18) . Hypertensive (n = 16) . . Pre . Post . Pre . Post . Pre . Post . Peak VO2 (mL/kg/minute) 16 ± 1 20 ± 1*** 16 ± 1 18 ± 1** 15 ± 1 19 ± 1*** Heart rate (bpm) 67 ± 3 65 ± 3 64 ± 2 62 ± 2 65 ± 2 62 ± 2 SBP (mmHg) 122 ± 5 114 ± 5 113 ± 4 124 ± 4* 138 ± 5 131 ± 4 DBP (mmHg) 68 ± 2 68 ± 4 63 ± 2 68 ± 3 77 ± 4 64 ± 3*** MBP (mmHg) 88 ± 4 82 ± 3 82 ± 2 85 ± 3 95 ± 4 82 ± 5* MSNA (burst/minute) 39 ± 3 33 ± 2** 48 ± 3 33 ± 2*** 45 ± 3 28 ± 2*** MSNA (burst/100 HB) 59 ± 5 52 ± 5* 75 ± 4 53 ± 3*** 75 ± 5 45 ± 3*** FBF (mL/minute/100 mL) 1.62 ± 0.09 1.90 ± 0.15* 1.63 ± 0.08 1.86 ± 0.12* 1.56 ± 0.11 2.16 ± 0.17*** FVC (a.u.) 1.85 ± 0.10 2.34 ± 0.20* 2.03 ± 0.12 2.27 ± 0.20 1.67 ± 0.11 2.55 ± 0.20*** . Idiopathic (n = 11) . Ischaemic (n = 18) . Hypertensive (n = 16) . . Pre . Post . Pre . Post . Pre . Post . Peak VO2 (mL/kg/minute) 16 ± 1 20 ± 1*** 16 ± 1 18 ± 1** 15 ± 1 19 ± 1*** Heart rate (bpm) 67 ± 3 65 ± 3 64 ± 2 62 ± 2 65 ± 2 62 ± 2 SBP (mmHg) 122 ± 5 114 ± 5 113 ± 4 124 ± 4* 138 ± 5 131 ± 4 DBP (mmHg) 68 ± 2 68 ± 4 63 ± 2 68 ± 3 77 ± 4 64 ± 3*** MBP (mmHg) 88 ± 4 82 ± 3 82 ± 2 85 ± 3 95 ± 4 82 ± 5* MSNA (burst/minute) 39 ± 3 33 ± 2** 48 ± 3 33 ± 2*** 45 ± 3 28 ± 2*** MSNA (burst/100 HB) 59 ± 5 52 ± 5* 75 ± 4 53 ± 3*** 75 ± 5 45 ± 3*** FBF (mL/minute/100 mL) 1.62 ± 0.09 1.90 ± 0.15* 1.63 ± 0.08 1.86 ± 0.12* 1.56 ± 0.11 2.16 ± 0.17*** FVC (a.u.) 1.85 ± 0.10 2.34 ± 0.20* 2.03 ± 0.12 2.27 ± 0.20 1.67 ± 0.11 2.55 ± 0.20*** Values are mean ± SE. Versus pre: *p < 0.05; **p < 0.01; ***p < 0.001. VO2: oxygen uptake; SBP: systolic blood pressure; DBP: diastolic blood pressure; MBP: mean blood pressure; MSNA: muscle sympathetic nerve activity; HB: heartbeats; FBF: forearm blood flow; FVC: forearm vascular conductance. Open in new tab Table 2. Effects of exercise training on physiologic parameters. . Idiopathic (n = 11) . Ischaemic (n = 18) . Hypertensive (n = 16) . . Pre . Post . Pre . Post . Pre . Post . Peak VO2 (mL/kg/minute) 16 ± 1 20 ± 1*** 16 ± 1 18 ± 1** 15 ± 1 19 ± 1*** Heart rate (bpm) 67 ± 3 65 ± 3 64 ± 2 62 ± 2 65 ± 2 62 ± 2 SBP (mmHg) 122 ± 5 114 ± 5 113 ± 4 124 ± 4* 138 ± 5 131 ± 4 DBP (mmHg) 68 ± 2 68 ± 4 63 ± 2 68 ± 3 77 ± 4 64 ± 3*** MBP (mmHg) 88 ± 4 82 ± 3 82 ± 2 85 ± 3 95 ± 4 82 ± 5* MSNA (burst/minute) 39 ± 3 33 ± 2** 48 ± 3 33 ± 2*** 45 ± 3 28 ± 2*** MSNA (burst/100 HB) 59 ± 5 52 ± 5* 75 ± 4 53 ± 3*** 75 ± 5 45 ± 3*** FBF (mL/minute/100 mL) 1.62 ± 0.09 1.90 ± 0.15* 1.63 ± 0.08 1.86 ± 0.12* 1.56 ± 0.11 2.16 ± 0.17*** FVC (a.u.) 1.85 ± 0.10 2.34 ± 0.20* 2.03 ± 0.12 2.27 ± 0.20 1.67 ± 0.11 2.55 ± 0.20*** . Idiopathic (n = 11) . Ischaemic (n = 18) . Hypertensive (n = 16) . . Pre . Post . Pre . Post . Pre . Post . Peak VO2 (mL/kg/minute) 16 ± 1 20 ± 1*** 16 ± 1 18 ± 1** 15 ± 1 19 ± 1*** Heart rate (bpm) 67 ± 3 65 ± 3 64 ± 2 62 ± 2 65 ± 2 62 ± 2 SBP (mmHg) 122 ± 5 114 ± 5 113 ± 4 124 ± 4* 138 ± 5 131 ± 4 DBP (mmHg) 68 ± 2 68 ± 4 63 ± 2 68 ± 3 77 ± 4 64 ± 3*** MBP (mmHg) 88 ± 4 82 ± 3 82 ± 2 85 ± 3 95 ± 4 82 ± 5* MSNA (burst/minute) 39 ± 3 33 ± 2** 48 ± 3 33 ± 2*** 45 ± 3 28 ± 2*** MSNA (burst/100 HB) 59 ± 5 52 ± 5* 75 ± 4 53 ± 3*** 75 ± 5 45 ± 3*** FBF (mL/minute/100 mL) 1.62 ± 0.09 1.90 ± 0.15* 1.63 ± 0.08 1.86 ± 0.12* 1.56 ± 0.11 2.16 ± 0.17*** FVC (a.u.) 1.85 ± 0.10 2.34 ± 0.20* 2.03 ± 0.12 2.27 ± 0.20 1.67 ± 0.11 2.55 ± 0.20*** Values are mean ± SE. Versus pre: *p < 0.05; **p < 0.01; ***p < 0.001. VO2: oxygen uptake; SBP: systolic blood pressure; DBP: diastolic blood pressure; MBP: mean blood pressure; MSNA: muscle sympathetic nerve activity; HB: heartbeats; FBF: forearm blood flow; FVC: forearm vascular conductance. Open in new tab Table 3. Relative changes (Δ%) in functional capacity and haemodynamic parameters. . Idiopathic (n = 11) . Ischaemic (n = 18) . Hypertensive (n = 16) . p–value . Peak VO2 (mL/kg/minute) 23 ± 2 16 ± 5 28 ± 7 0.34 Heart rate (bpm) −3 ± 3 −2 ± 3 −4 ± 2 0.94 SBP (mmHg) −6 ± 3 11 ± 3a,b −5 ± 3 0.003 DBP (mmHg) 0 ± 7 8 ± 4 −16 ± 3c 0.008 MBP (mmHg) −6 ± 3 5 ± 4 −11 ± 6 0.08 . Idiopathic (n = 11) . Ischaemic (n = 18) . Hypertensive (n = 16) . p–value . Peak VO2 (mL/kg/minute) 23 ± 2 16 ± 5 28 ± 7 0.34 Heart rate (bpm) −3 ± 3 −2 ± 3 −4 ± 2 0.94 SBP (mmHg) −6 ± 3 11 ± 3a,b −5 ± 3 0.003 DBP (mmHg) 0 ± 7 8 ± 4 −16 ± 3c 0.008 MBP (mmHg) −6 ± 3 5 ± 4 −11 ± 6 0.08 Values are mean ± SE. a Versus idiopathic. b Versus hypertensive. c Versus ischaemic. VO2: oxygen uptake; SBP: systolic blood pressure; DBP: diastolic blood pressure; MBP: mean blood pressure. Open in new tab Table 3. Relative changes (Δ%) in functional capacity and haemodynamic parameters. . Idiopathic (n = 11) . Ischaemic (n = 18) . Hypertensive (n = 16) . p–value . Peak VO2 (mL/kg/minute) 23 ± 2 16 ± 5 28 ± 7 0.34 Heart rate (bpm) −3 ± 3 −2 ± 3 −4 ± 2 0.94 SBP (mmHg) −6 ± 3 11 ± 3a,b −5 ± 3 0.003 DBP (mmHg) 0 ± 7 8 ± 4 −16 ± 3c 0.008 MBP (mmHg) −6 ± 3 5 ± 4 −11 ± 6 0.08 . Idiopathic (n = 11) . Ischaemic (n = 18) . Hypertensive (n = 16) . p–value . Peak VO2 (mL/kg/minute) 23 ± 2 16 ± 5 28 ± 7 0.34 Heart rate (bpm) −3 ± 3 −2 ± 3 −4 ± 2 0.94 SBP (mmHg) −6 ± 3 11 ± 3a,b −5 ± 3 0.003 DBP (mmHg) 0 ± 7 8 ± 4 −16 ± 3c 0.008 MBP (mmHg) −6 ± 3 5 ± 4 −11 ± 6 0.08 Values are mean ± SE. a Versus idiopathic. b Versus hypertensive. c Versus ischaemic. VO2: oxygen uptake; SBP: systolic blood pressure; DBP: diastolic blood pressure; MBP: mean blood pressure. Open in new tab Exercise training provoked a significant reduction in MSNA (burst/minute) and MSNA (burst/100 HB) in idiopathic (p = 0.003 and p = 0.02, respectively), ischaemic (p < 0.001 and p < 0.001, respectively) and hypertensive (p < 0.001 and p < 0.001, respectively) patients (Table 2). The comparisons between groups showed that the reductions in MSNA (burst/minute) and MSNA (burst/100 HB) were greater in hypertensive patients compared with those in idiopathic patients (Figure 1). No differences were observed between hypertensive and ischaemic patients (MSNA frequency: −34% vs. −30%, respectively; MSNA incidence: −31% vs. −27%, respectively). There was a tendency towards a difference between ischaemic and idiopathic patients in MSNA frequency (p = 0.07) (Figure 1). Figure 1. Open in new tabDownload slide Muscle sympathetic nerve activity (MSNA). (a) Relative changes (Δ%) in MSNA frequency and (b) MSNA incidence in exercise-trained idiopathic, ischaemic and hypertensive heart failure patients. *Versus idiopathic heart failure patients. Exercise training significantly increased FBF in idiopathic (p = 0.02), ischaemic (p = 0.02) and hypertensive (p < 0.001) patients (Table 2). Exercise training significantly increased FVC in idiopathic (p = 0.01) and hypertensive (p < 0.001) patients, but caused no changes in ischaemic patients (p = 0.09) (Table 2). The increase in FBF was higher in hypertensive patients than in idiopathic and ischaemic patients (42% vs. 17% and 15%, respectively) (Figure 2(a)). The changes in FBF were similar in idiopathic and ischaemic patients (Figure 2(a)). Exercise training caused a greater increase in FVC in hypertensive patients than in idiopathic and ischaemic patients (57% vs. 26% and 13%, respectively) (Figure 2(b)). The changes were similar in idiopathic and ischaemic groups (Figure 2(b)). Figure 2. Open in new tabDownload slide Forearm blood flow (FBF) and forearm vascular conductance. (a) Relative changes (Δ%) in FBF and (b) forearm vascular conductance (FVC) in exercise-trained idiopathic, ischaemic and hypertensive heart failure patients. *Versus idiopathic heart failure patients. †Versus ischaemic heart failure patients. Further analyses showed that the reduction in MSNA (burst/minute) was significantly correlated with the increases in FBF (p = 0.04 and r = −0.50) and FVC (p = 0.04 and r = −0.51) in the hypertensive group. No correlations between MSNA and FBF or FVC were found in the ischaemic and idiopathic groups. The increase in FBF was significantly correlated with the increase in peak VO2 (p = 0.04 and r = 0.50) in the ischaemic group. No correlations between FBF or FVC and peak VO2 were found in the idiopathic and hypertensive groups. Discussion The main and new findings of the present study are that after exercise training: (1) the reduction in MSNA is more pronounced in hypertensive HF patients than in idiopathic HF patients; (2) the increase in FBF is greater in hypertensive HF patients than in idiopathic and ischaemic HF patients; and (3) the increase in peak VO2 is similar in idiopathic, ischaemic and hypertensive HF patients. Sympathetic activation is the hallmark of HF. This autonomic abnormality is directly associated with the clinical severity and prognosis of HF patients.17,18 This sympathetic alteration is present in all aetiologies, even though controversy remains regarding the level of sympathetic activation across sets of patients. Notarius et al. reported that ischaemic HF patients had greater MSNA than non-ischaemic HF patients.19 In contrast, Grassi et al. found no difference in MSNA levels in ischaemic and non-ischaemic HF patients.20 We further observed that MSNA levels were similar among idiopathic, ischaemic and chagasic HF patients.21 The present study confirms these findings. Baseline MSNA levels were similar in idiopathic, ischaemic and hypertensive patients. Exercise training has been shown to dramatically reduce MSNA levels in chronic HF patients.7 This benefit of exercise has been observed regardless of β-blocker therapy, age and sex.9,11,12 In a recent study, we found that not only voluntary muscle contraction during conventional exercise training, but also involuntary muscle contraction by functional electrical stimulation caused substantial reductions in MSNA levels in patients admitted to hospital for stabilization of HF.22 Taken together, these findings confer to muscle contraction a unique strategy for decreasing sympathetic activity in HF patients. The present study extends the knowledge that the reduction in MSNA after exercise training is greater in hypertensive HF patients. Our study provides no explanation for the decrease of MSNA levels in hypertensive, idiopathic and ischaemic HF patients. However, some might suggest that the arterial baroreflex control contributes to this autonomic response. Enhanced arterial baroreflex control of renal sympathetic nerve activity,23 which is associated with increased afferent aortic depressor nerve sensitivity,24 has been described in an exercise-trained animal model of HF. In humans, exercise training improves arterial baroreflex sensitivity in HF patients.25 More recently, Groehs et al. reported that exercise training prevented deterioration in the arterial baroreflex control of MSNA in HF patients.26 The decrease in MSNA after exercise training may be due, in part, to a reduction in chemoreflex hypersensitivity. Some investigators have elegantly reported that exercise training normalized the peripheral chemoreflex control of renal sympathetic nerve activity in an animal model of HF.27 The deactivation in muscle afferent mechanoreflex sensitivity can also be explained by a reduction in sympathetic outflow.13 In a recent study, we demonstrated that exercise training provoked a significant decrease in mechanoreflex control of MSNA in HF patients.13 The mechanisms involved in the more marked reduction in MSNA in hypertensive HF patients remain unknown. In fact, this is an interesting topic for future investigations. Peripheral vasoconstriction status has been consistently reported in patients with chronic HF. Moreover, studies regarding vascular function provide information that the reduction in FBF is due to both augmented sympathetic outflow and impaired endothelial function.28–30 Intra-arterial infusion of phentolamine significantly increases FBF in HF patients.29 However, intra-arterial infusion of nitric oxide synthase inhibitor NG-Monomethyl-L-Arginine (L-NMMA) does not change FBF.28 The present study confirms that exercise training significantly increases FBF and FVC in HF patients. This is an important finding because some medications for the treatment of HF cause no changes in muscle blood flow. For instance, carvedilol administration for 6 months provoked no changes in FBF in HF patients.31 The novelty of the present study is that the changes in FBF are more marked in hypertensive patients than in idiopathic and ischaemic patients. The mechanisms underlying this finding are beyond the scope of our study. However, they may include a reduction in MSNA. In fact, in the set of hypertensive patients, significant correlations between the changes in MSNA and the changes in FBF (p = 0.04 and r = −0.50) and FVC (p = 0.04 and r = −0.51) were found. Of course, we cannot rule out the possibility that the improvement in endothelial function is greater in hypertensive patients than in the other two sets of patients. Whether or not this interpretation is correct, the less pronounced changes in FBF levels in exercise-trained ischaemic and idiopathic patients can be attributed to lower changes in endothelial function and MSNA levels. A previous study showed that the flow-mediated dilatation in the brachial artery was impaired in ischaemic HF patients compared with that in healthy subjects.32 This response was not seen in non-ischaemic HF patients. In addition, MSNA reduction was less pronounced in ischaemic and idiopathic HF patients in our study.32 Because the changes in MSNA and FBF were different in hypertensive, idiopathic and ischaemic HF patients, some might expect that these changes would be translated to the peak VO2 across these sets of patients. Surprisingly, this was not the case. The increases in peak VO2 were similar among groups. One possible explanation for this finding is that functional capacity is not only dependent on neurovascular changes, but also on skeletal muscle changes. In fact, increases in exercise tolerance associated with changes in skeletal muscle have been extensively reported in chronic HF patients. Exercise training has been shown to reduce pro-inflammatory responses and protein degradation and to increase the cross-sectional area and calcium cycling in skeletal muscle.33–37 Alternatively, the contribution of the neurovascular force and skeletal muscle force to the improvement in functional capacity may be different in exercise-trained hypertensive, idiopathic and ischaemic patients. We found that the increase in FBF was directly correlated with the increase in functional capacity in ischaemic patients (p = 0.04 and r = 0.50), but not in idiopathic and hypertensive patients. Future studies should focus on the effects of exercise training on skeletal muscle in HF patients with different aetiologies. Our study has clinical implications. Both MSNA and FBF are independent predictors of mortality in HF.18 Thus, the reduction in MSNA and the increase in FBF may contribute to patient outcomes, especially in hypertensive patients. Functional capacity is inversely associated with mortality rate.38 The present study provides evidence that exercise training improves functional capacity regardless of aetiology. We recognize the limitations in our study. The distribution of the sexes was different across aetiologies. It is unlikely that this argument limits our study, since we previously observed that the benefits of exercise training on neurovascular control and functional capacity are independent of sex in HF patients.11 The use of statins was greater in ischaemic patients than in hypertensive and idiopathic patients. Some, but not all, propose that statins reduce MSNA.39,40 Future studies should focus on this issue. We cannot rule out the possibility that the small number of patients included in our study and the retrospective design from a single centre may have influenced the present findings regarding the benefits of exercise training on neurovascular control in different sets of HF patients. In conclusion, the present study provides evidence that the magnitude of exercise training induced a reduction in MSNA, and an increase in FBF depends on the HF aetiology. The increase in functional capacity is independent of aetiology. Author contribution LMA-C, MJNNA, MUPBR and CEN contributed to the conception or design of the work. All authors contributed to the acquisition, analysis or interpretation of the data for the work. LMA-C, LMU-P, PFT, MJNNA and MUPBR drafted the manuscript. LMA-C, LMU-P, MRS, MUPBR and CEN critically revised the manuscript. All authors gave final approval and agree to be accountable for all aspects of the work, ensuring its integrity and accuracy. 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) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: this work was supported by FAPESP (#2010/50048-1) and Fundação Zerbini. LMA-C (#2013/15651-7) and LMU-P (#03/10881-2) were supported by FAPESP. LMA-C (#142366/2009-9), MRS (#234052/2014-7), MUPBR (#309821/2014-2) and CEN (#303573/2015-5) were supported by CNPq. MRS was supported CAPES. References 1 Roger VL . Epidemiology of heart failure . Circ Res 2013 ; 113 : 646 – 659 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Negrao CE , Middlekauff HR, Gomes-Santos ILet al. Effects of exercise training on neurovascular control and skeletal myopathy in systolic heart failure . Am J Physiol Heart Circ Physiol 2015 ; 308 : H792 – H802 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Gheorghiade M , Bonow RO. Chronic heart failure in the United States: A manifestation of coronary artery disease . Circulation 1998 ; 97 : 282 – 289 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Roger VL . The heart failure epidemic . Int J Environ Res Public Health 2010 ; 7 : 1807 – 1830 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Arena R , Myers J, Abella Jet al. Influence of heart failure etiology on the prognostic value of peak oxygen consumption and minute ventilation/carbon dioxide production slope . Chest 2005 ; 128 : 2812 – 2817 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Yancy CW , Jessup M, Bozkurt Bet al. 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines . J Am Coll Cardiol 2013 ; 62 : e147 – e239 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Roveda F , Middlekauff HR, Rondon MUet al. The effects of exercise training on sympathetic neural activation in advanced heart failure: A randomized controlled trial . J Am Coll Cardiol 2003 ; 42 : 854 – 860 . Google Scholar Crossref Search ADS PubMed WorldCat 8 de Mello Franco FG , Santos AC, Rondon MUet al. Effects of home-based exercise training on neurovascular control in patients with heart failure . Eur J Heart Fail 2006 ; 8 : 851 – 855 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Fraga R , Franco FG, Roveda Fet al. Exercise training reduces sympathetic nerve activity in heart failure patients treated with carvedilol . Eur J Heart Fail 2007 ; 9 : 630 – 636 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Ueno LM , Drager LF, Rodrigues ACet al. Effects of exercise training in patients with chronic heart failure and sleep apnea . Sleep 2009 ; 32 : 637 – 647 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Antunes-Correa LM , Melo RC, Nobre TSet al. Impact of gender on benefits of exercise training on sympathetic nerve activity and muscle blood flow in heart failure . Eur J Heart Fail 2010 ; 12 : 58 – 65 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Antunes-Correa LM , Kanamura BY, Melo RCet al. Exercise training improves neurovascular control and functional capacity in heart failure patients regardless of age . Eur J Prev Cardiol 2012 ; 19 : 822 – 829 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Antunes-Correa LM , Nobre TS, Groehs RVet al. Molecular basis for the improvement in muscle metaboreflex and mechanoreflex control in exercise-trained humans with chronic heart failure . Am J Physiol Heart Circ Physiol 2014 ; 307 : H1655 – H1666 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Piña IL , Bittner V, Clare RMet al. Effects of exercise training on outcomes in women with heart failure: Analysis of HF-ACTION (Heart Failure – A Controlled Trial Investigating Outcomes of Exercise TraiNing) by sex . JACC Heart Fail 2014 ; 2 : 180 – 186 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Flynn KE , Piña IL, Whellan DJet al. Effects of exercise training on health status in patients with chronic heart failure: HF-ACTION randomized controlled trial . JAMA 2009 ; 301 : 1451 – 1459 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Sagar VA , Davies EJ, Briscoe Set al. Exercise-based rehabilitation for heart failure: Systematic review and meta-analysis . Open Heart 2015 ; 2 : e000163 – e000163 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Negrao CE , Middlekauff HR. Adaptations in autonomic function during exercise training in heart failure . Heart Fail Rev 2008 ; 13 : 51 – 60 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Barretto AC , Santos AC, Munhoz Ret al. Increased muscle sympathetic nerve activity predicts mortality in heart failure patients . Int J Cardiol 2009 ; 135 : 302 – 307 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Notarius CF , Spaak J, Morris BLet al. Comparison of muscle sympathetic activity in ischemic and nonischemic heart failure . J Card Fail 2007 ; 13 : 470 – 475 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Grassi G , Seravalle G, Bertinieri Get al. Sympathetic and reflex abnormalities in heart failure secondary to ischaemic or idiopathic dilated cardiomyopathy . Clin Sci (Lond) 2001 ; 101 : 141 – 146 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Negrão CE , Santos AC, Rondon MUet al. Muscle sympathetic nerve activity in patients with Chagas’ disease . Int J Cardiol 2009 ; 137 : 252 – 259 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Groehs RV , Antunes-Correa LM, Nobre TSet al. Muscle electrical stimulation improves neurovascular control and exercise tolerance in hospitalised advanced heart failure patients . Eur J Prev Cardiol 2016 ; 23 : 1599 – 1608 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Mousa TM , Liu D, Cornish KGet al. Exercise training enhances baroreflex sensitivity by an angiotensin II-dependent mechanism in chronic heart failure . J Appl Physiol 1985 2008 ; 104 : 616 – 624 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 24 Rondon E , Brasileiro-Santos MS, Moreira EDet al. Exercise training improves aortic depressor nerve sensitivity in rats with ischemia-induced heart failure . Am J Physiol Heart Circ Physiol 2006 ; 291 : H2801 – H2806 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Iellamo F , Manzi V, Caminiti Get al. Dose–response relationship of baroreflex sensitivity and heart rate variability to individually-tailored exercise training in patients with heart failure . Int J Cardiol 2013 ; 166 : 334 – 339 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Groehs RV , Toschi-Dias E, Antunes-Correa LMet al. Exercise training prevents the deterioration in the arterial baroreflex control of sympathetic nerve activity in chronic heart failure patients . Am J Physiol Heart Circ Physiol 2015 ; 308 : H1096 – H1102 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Li YL , Ding Y, Agnew Cet al. Exercise training improves peripheral chemoreflex function in heart failure rabbits . J Appl Physiol 2008 ; 105 : 782 – 790 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Nazaré Nunes Alves MJ , Alves MJ, dos Santos MRet al. Mechanisms of blunted muscle vasodilation during peripheral chemoreceptor stimulation in heart failure patients . Hypertension 2012 ; 60 : 669 – 676 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Alves MJ , Rondon MU, Santos ACet al. Sympathetic nerve activity restrains reflex vasodilatation in heart failure . Clin Auton Res 2007 ; 17 : 364 – 369 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Sandri M , Viehmann M, Adams Vet al. Chronic heart failure and aging – Effects of exercise training on endothelial function and mechanisms of endothelial regeneration: Results from the Leipzig Exercise Intervention in Chronic heart failure and Aging (LEICA) study . Eur J Prev Cardiol 2016 ; 23 : 349 – 358 . Google Scholar Crossref Search ADS PubMed WorldCat 31 De Matos LD , Gardenghi G, Rondon MUet al. Impact of 6 months of therapy with carvedilol on muscle sympathetic nerve activity in heart failure patients . J Card Fail 2004 ; 10 : 496 – 502 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Shah A , Gkaliagkousi E, Ritter JMet al. Endothelial function and arterial compliance are not impaired in subjects with heart failure of non-ischemic origin . J Card Fail 2010 ; 16 : 114 – 120 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Gielen S , Adams V, Möbius-Winkler Set al. Anti-inflammatory effects of exercise training in the skeletal muscle of patients with chronic heart failure . J Am Coll Cardiol 2003 ; 42 : 861 – 868 . Google Scholar Crossref Search ADS PubMed WorldCat 34 Dos Santos MR , Sayegh AL, Bacurau AVet al. Effect of exercise training and testosterone replacement on skeletal muscle wasting in patients with heart failure with testosterone deficiency . Mayo Clin Proc 2016 ; 91 : 575 – 586 . Google Scholar Crossref Search ADS PubMed WorldCat 35 Bacurau AV , Jardim MA, Ferreira JCet al. Sympathetic hyperactivity differentially affects skeletal muscle mass in developing heart failure: Role of exercise training . J Appl Physiol (1985) 2009 ; 106 : 1631 – 1640 . Google Scholar Crossref Search ADS PubMed WorldCat 36 Bueno CR , Ferreira JC, Pereira MGet al. Aerobic exercise training improves skeletal muscle function and Ca2+ handling-related protein expression in sympathetic hyperactivity-induced heart failure . J Appl Physiol 1985 2010 ; 109 : 702 – 709 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 37 Nobre TS , Antunes-Correa LM, Groehs RVet al. Exercise training improves neurovascular control and calcium cycling gene expression in heart failure patients with cardiac resynchronization therapy . Am J Physiol Heart Circ Physiol 2016 ; 311 : H1180 – H1188 . Google Scholar Crossref Search ADS PubMed WorldCat 38 Chase PJ , Kenjale A, Cahalin LPet al. Effects of respiratory exchange ratio on the prognostic value of peak oxygen consumption and ventilatory efficiency in patients with systolic heart failure . JACC Heart Fail 2013 ; 1 : 427 – 432 . Google Scholar Crossref Search ADS PubMed WorldCat 39 Horwich TB , Middlekauff HR. Potential autonomic nervous system effects of statins in heart failure . Heart Fail Clin 2008 ; 4 : 163 – 167 . Google Scholar Crossref Search ADS PubMed WorldCat 40 Horwich TB , Middlekauff HR, Maclellan WRet al. Statins do not significantly affect muscle sympathetic nerve activity in humans with nonischemic heart failure: A double-blind placebo-controlled trial . J Card Fail 2011 ; 17 : 879 – 886 . Google Scholar Crossref Search ADS PubMed WorldCat © The European Society of Cardiology 2017 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 2017

Abstract Background Despite atrial fibrillation representing an established risk factor for stroke, the association between atrial fibrillation and both progression of coronary atherosclerosis and major adverse cardiovascular events is not well characterized. We assessed the serial measures of coronary atheroma burden and cardiovascular outcomes in patients with and without atrial fibrillation. Methods Data were analyzed from nine clinical trials involving 4966 patients with coronary artery disease undergoing serial intravascular ultrasonography at 18–24 month intervals to assess changes in percent atheroma volume (PAV). Using a propensity weighted analysis, and following adjustment for baseline variables, patients with (n = 190) or without (n = 4776) atrial fibrillation were compared with regard to coronary plaque volume and major adverse cardiovascular events (death, myocardial infarction, and stroke). Results Atrial fibrillation patients demonstrated lower baseline PAV (36.0 ± 8.9 vs. 38.1 ± 8.9%, p = 0.002) and less PAV progression (–0.07 ± 0.34 vs. + 0.23 ± 0.34%, p = 0.001) compared with the non-atrial fibrillation group. Multivariable analysis revealed atrial fibrillation to independently predict both myocardial infarction [HR, 2.41 (1.74,3.35), p<0.001] 2.41 (1.74, 3.35), p < 0.00) and major adverse cardiovascular events [HR, 2.2, (1.66, 2.92), p<0.001] 2.20 (1.66, 2.92), p < 0.001]. Kaplan–Meier analysis showed that atrial fibrillation compared with non-atrial fibrillation patients had a significantly higher two-year cumulative incidence of overall major adverse cardiovascular events (4.4 vs. 2.0%, log-rank p = 0.02) and myocardial infarction (3.3 vs. 1.5%, log-rank p = 0.05). Conclusions The presence of atrial fibrillation independently associates with a heightened risk of myocardial infarction despite a lower baseline burden and progression rate of coronary atheroma. Further studies are necessary to define the pathogenesis of myocardial infarction in the setting of atrial fibrillation. Intravascular ultrasound, atrial fibrillation, myocardial infarction Introduction Atrial fibrillation is a well-established risk factor for athero-thrombotic vascular events. Atrial fibrillation affects 1–2% of the general population1 and is associated with a five-fold increased risk of stroke, two-fold overall increased risk of death and coronary artery disease (CAD).1 Ischemic heart disease is the second most common comorbid chronic condition among Medicare beneficiaries with atrial fibrillation.2 Epidemiologic studies demonstrated that ischemic heart disease and diabetes3 are risk factors for developing atrial fibrillation,4,5 and the Framingham Study reported that atrial fibrillation patients frequently harbor the same cardiovascular risk factors for myocardial infarction (MI).4 Conversely, prospective cohort studies suggest that atrial fibrillation could also serve as a risk factor for future MI.6-9 The factors that underscore this potential link remain poorly characterized. Intravascular ultrasound (IVUS), by virtue of its ability to precisely measure coronary atheroma volume, permits characterization of factors associating with plaque progression. To date, no serial coronary artery imaging studies have investigated the impact of atrial fibrillation upon the course of coronary atherosclerosis and subsequent cardiovascular outcomes. The current analysis tested the hypothesis that atrial fibrillation patients would exhibit accelerated coronary atheroma progression compared with their non-atrial fibrillation counterparts, which would translate into a greater risk for incident MI. Methods Subject selection and study design The current analysis pooled data from nine prospective serial coronary IVUS imaging trials, including 4966 patients with established CAD. Included in this analysis were trials assessing intensive lipid lowering with statins (REVERSAL (Reversal of Atherosclerosis With Aggressive Lipid Lowering), ASTEROID (A Study to Evaluate the Effect of Rouvastatin on Intravascular-Ultrasound Derived Indices of Coronary Atheroma Burden), and SATURN (The Study of Coronary Atheroma by Intravascular Ultrasound: Effect of Rosuvastatin Versus Atorvastatin),10-12 anti-hypertensive therapies (AQUARIUS (Aliskiren Quantitative Atherosclerosis Regression Intravascular Ultrasound Study) and NORMALIZE (Norvasc for Regression of Manifest Atherosclerotic Lesions by Intravascular Sonographic Evaluation),13,14 the anti-atherosclerotic efficacy of acyl-coenzyme A:cholesteryl ester transfer protein inhibition (ACTIVATE (ACAT Intravascular Atherosclerosis Treatment Evaluation),15 cholesteryl ester transfer protein inhibition (ILLUSTRATE (Investigation of Lipid Level Management Using Coronary Ultrasound to Assess Reduction of Atherosclerosis by CETP Inhibition and HDL Elevation),16 endocannabinoid receptor antagonism (STRADIVARIUS (Strategy to Reduce Atherosclerosis Development Involving Administration of Rimonabont – The Intravascular Ultrasound Study)),17 and the peroxisome proliferator-activated receptor gamma (PERISCOPE (Comparison of Pioglitazone versus Glimepiride on Progression of Coronary Atherosclerosis in Patients With Type 2 Diabetes).18 All patients were required to have CAD, defined as having at least one lumen narrowing >20% within a major epicardial coronary artery during a clinically indicated coronary angiogram. Acquisition and analysis of IVUS images The acquisition and serial analysis of IVUS images in each of these trials has been previously described in detail.10-18 Briefly, following anticoagulation and intracoronary nitroglycerin administration, an imaging catheter containing a high-frequency ultrasound transducer (30–45 MHz) was inserted distally within a coronary artery. Target vessels for imaging were selected if they contained no luminal stenosis>50% angiographic severity within a segment of at least 30 mm length, had not undergone previous revascularization, and were not considered to be the culprit vessel for a prior MI. Continuous ultrasonic imaging was acquired during withdrawal of the catheter through the segment of artery at a constant rate of 0.5 mm/s. Imaging was performed within the same coronary artery at baseline and at study completion, which ranged from 18 to 24 months. Imaging in all trials was screened by the Atherosclerosis Imaging Core Laboratory of the Cleveland Clinic Coordinating Center for Clinical Research. Patients meeting pre-specified requirements for image quality were eligible for randomization. An anatomically matched segment was defined at the two time points on the basis of proximal and distal side branches (fiduciary points). Cross-sectional images spaced precisely 1 mm apart were selected for measurement. Leading edges of the lumen and external elastic membrane (EEM) were traced by manual planimetry. Plaque area was defined as the area occupied between these leading edges. The percent atheroma volume (PAV) was calculated as the proportion of vessel wall volume occupied by atherosclerotic plaque: Open in new tabDownload slide The total atheroma volume (TAV) was calculated by summation of the plaque area calculated for each measured image and subsequently normalized to account for differences in segment length between subjects: TAVNormalizd=∑(EEMarea-Lumenarea)Numberofimagesinpullback×mediannumberofimagesincohort Volumes occupied by the lumen and EEM were similarly calculated by summation of their respective areas in each measured image and subsequently normalized to account for differences in segment length between subjects. Definition of atrial fibrillation and MI Atrial fibrillation was identified by a history of physician diagnosis of atrial fibrillation reported by patients during the baseline visit. Major adverse cardiovascular events (MACEs: death, MI, stroke) were reported by physicians in the setting of randomized clinical trials, and were formally adjudicated in two IVUS trials (SATURN and AQUARIUS). Statistical analyses Continuous variables were reported as mean ± SD if normally distributed and as median (interquartile range) if non-normally distributed. Demographics, baseline clinical characteristics and medications, laboratory biochemical data, and baseline IVUS parameters were summarized and stratified by baseline atrial fibrillation status, and were compared using two-sample t-test for normally distributed continuous variables, Wilcoxon rank–sum test for non-normally distributed continuous variables, and chi-square test (or exact test) for categorical variables. Due to significant differences in various baseline characteristics between the groups and the small sample size of the atrial fibrillation group, a propensity score weighting method was applied. The propensity scores and corresponding inverse probability of treatment weight (IPTW) were calculated through a logistic regression model with baseline atrial fibrillation status as the outcome and using all the related baseline characteristics and medications as covariates. Improvement of the balance of the baseline covariates was assessed by comparing standardized differences of these covariates before and after IPTW. All subsequent analyses were weighted by IPTW and adjusted for clinical trial. Serial changes in IVUS measurements (reported as least-squares mean ± SE) were analyzed by linear mixed models adjusting for their baseline counterparts and treating trial as a random effect. A full multivariable linear regression model was further constructed to identify factors associated with change in PAV. Furthermore, association between atrial fibrillation and MACE was explored via survival analysis. Kaplan–Meier plots were generated to compare the cumulative survival rates between atrial fibrillation and non-atrial fibrillation. Cox proportional hazards regression models with or without covariate adjustment were created for MACE and MI respectively. In these multivariable models, covariates were selected by bootstrapping (1000 iterations and a p-value criterion of 0.05 for retention). Those variables having a 40% or more probability of retention were entered into a second model with the stepwise model selection procedure. The significance level to enter and keep a variable was set at 0.05. For change in PAV, linear spline terms were created and applied in the Cox models. The selected covariates form the covariate set for the final multivariable models. A two-sided p-value of < 0.05 was considered statistically significant. All analyses were conducted using the SAS software version 9.2 (SAS Institute, Cary, North Carolina, USA). Results Clinical and biochemical characteristics Table 1 summarizes baseline patient characteristics. Compared with non-atrial fibrillation patients (n = 4776), atrial fibrillation patients (n = 190) were more likely to be older (62.9 ± 9.0 vs. 57.7 ± 9.2 years, p < 0.001), have higher body mass index (31.7 ± 5.7 vs. 30.7 ± 5.8 kg/m2), higher rates of hypertension (86.8 vs. 77.3%, p = 0.002), and prior coronary artery bypass grafting (6.3 vs. 2.2%, p = 0.003), more likely to be receiving beta-blockers (82.1 vs. 74.1%, p < 0.01), yet less likely to be treated with statins (68.4 vs. 74.6%, p = 0.06) and aspirin (84.2 vs. 93.7%, p < 0.001). The degree of risk factor control at baseline and follow-up are summarized in Table 2. No significant differences were noted between groups with regard to baseline and follow-up risk factor control. Table 1. Clinical characteristics and use of established medical therapies in AF and non-AF patients. Characteristic . AF (n = 190) . Non-AF (n = 4776) . p-value . Age, years 62.9 ± 9.03 57.7 ± 9.15 <0.001 Female, % 23.7 28.3 0.17 Hypertension, % 86.8 77.3 0.002 Current smoker, % 21.6 25.1 0.29 BMI, kg/m2 31.7 ± 5.7 30.7 ± 5.8 0.01 Diabetes, % 23.7 29.2 0.10 Metabolic syndrome, % 61.1 54.4 0.07 History of MI, % 27.9 29.1 0.71 History of CHF, % 15.3 2.9 <0.001 History of CABG, % 6.3 2.2 0.003 History of PCI, % 36.9 39.6 0.48 ACS (at presentation) 22.2 25.3 0.35 Statin use, %  Baseline 68.4 74.6 0.06  Concomitant 92.1 95.6 0.02 Beta-blocker use, %  Baseline 82.1 74.1 0.01  Concomitant 84.7 75.8 0.004 ACE/ARB use, %  Baseline 48.5 51.0 0.53  Concomitant 58.0 55.1 0.47 Aspirin use, %  Baseline 84.2 93.7 <0.001  Concomitant 85.8 94.5 <0.001 Characteristic . AF (n = 190) . Non-AF (n = 4776) . p-value . Age, years 62.9 ± 9.03 57.7 ± 9.15 <0.001 Female, % 23.7 28.3 0.17 Hypertension, % 86.8 77.3 0.002 Current smoker, % 21.6 25.1 0.29 BMI, kg/m2 31.7 ± 5.7 30.7 ± 5.8 0.01 Diabetes, % 23.7 29.2 0.10 Metabolic syndrome, % 61.1 54.4 0.07 History of MI, % 27.9 29.1 0.71 History of CHF, % 15.3 2.9 <0.001 History of CABG, % 6.3 2.2 0.003 History of PCI, % 36.9 39.6 0.48 ACS (at presentation) 22.2 25.3 0.35 Statin use, %  Baseline 68.4 74.6 0.06  Concomitant 92.1 95.6 0.02 Beta-blocker use, %  Baseline 82.1 74.1 0.01  Concomitant 84.7 75.8 0.004 ACE/ARB use, %  Baseline 48.5 51.0 0.53  Concomitant 58.0 55.1 0.47 Aspirin use, %  Baseline 84.2 93.7 <0.001  Concomitant 85.8 94.5 <0.001 AF: atrial fibrillation; ACE: angiotensin-converting enzyme; ACS: acute coronary syndrome; ARB: angiotensin receptor blocker; BMI: body mass index; CABG: coronary artery bypass grafting; MI: myocardial infarction; PCI: percutaneous coronary intervention. Open in new tab Table 1. Clinical characteristics and use of established medical therapies in AF and non-AF patients. Characteristic . AF (n = 190) . Non-AF (n = 4776) . p-value . Age, years 62.9 ± 9.03 57.7 ± 9.15 <0.001 Female, % 23.7 28.3 0.17 Hypertension, % 86.8 77.3 0.002 Current smoker, % 21.6 25.1 0.29 BMI, kg/m2 31.7 ± 5.7 30.7 ± 5.8 0.01 Diabetes, % 23.7 29.2 0.10 Metabolic syndrome, % 61.1 54.4 0.07 History of MI, % 27.9 29.1 0.71 History of CHF, % 15.3 2.9 <0.001 History of CABG, % 6.3 2.2 0.003 History of PCI, % 36.9 39.6 0.48 ACS (at presentation) 22.2 25.3 0.35 Statin use, %  Baseline 68.4 74.6 0.06  Concomitant 92.1 95.6 0.02 Beta-blocker use, %  Baseline 82.1 74.1 0.01  Concomitant 84.7 75.8 0.004 ACE/ARB use, %  Baseline 48.5 51.0 0.53  Concomitant 58.0 55.1 0.47 Aspirin use, %  Baseline 84.2 93.7 <0.001  Concomitant 85.8 94.5 <0.001 Characteristic . AF (n = 190) . Non-AF (n = 4776) . p-value . Age, years 62.9 ± 9.03 57.7 ± 9.15 <0.001 Female, % 23.7 28.3 0.17 Hypertension, % 86.8 77.3 0.002 Current smoker, % 21.6 25.1 0.29 BMI, kg/m2 31.7 ± 5.7 30.7 ± 5.8 0.01 Diabetes, % 23.7 29.2 0.10 Metabolic syndrome, % 61.1 54.4 0.07 History of MI, % 27.9 29.1 0.71 History of CHF, % 15.3 2.9 <0.001 History of CABG, % 6.3 2.2 0.003 History of PCI, % 36.9 39.6 0.48 ACS (at presentation) 22.2 25.3 0.35 Statin use, %  Baseline 68.4 74.6 0.06  Concomitant 92.1 95.6 0.02 Beta-blocker use, %  Baseline 82.1 74.1 0.01  Concomitant 84.7 75.8 0.004 ACE/ARB use, %  Baseline 48.5 51.0 0.53  Concomitant 58.0 55.1 0.47 Aspirin use, %  Baseline 84.2 93.7 <0.001  Concomitant 85.8 94.5 <0.001 AF: atrial fibrillation; ACE: angiotensin-converting enzyme; ACS: acute coronary syndrome; ARB: angiotensin receptor blocker; BMI: body mass index; CABG: coronary artery bypass grafting; MI: myocardial infarction; PCI: percutaneous coronary intervention. Open in new tab Table 2. Measures of risk factor control in AF and non-AF patients. . AF . Non-AF . p-value . Triglycerides, mg/dl  Baseline 139.0 (97.3, 211.3) 137.0 (97.8, 193.0) 0.66  Follow-up 131.8 (99.7, 184.0) 127.1 (94.2, 171.8) 0.24  Change –16.0 ± 70.5 –12.4 ± 70.07 0.65 HDL-C, mg/dl  Baseline 42.2 ± 11.5 43.4 ± 11.8 0.24  Follow-up 48.0 ± 15.1 48.2 ± 14.7 0.73  Change 5.8 ± 10.8 4.9 ± 10.2 0.68 Systolic blood pressure, mm/Hg  Baseline 127.0 ± 15.43 128.5 ± 16.24 0.34  Follow-up 130.3 ± 15.1 129.4 ± 13.7 0.24  Change 3.3 ± 16.01 0.9 ± 15.4 0.08 Diastolic blood pressure, mm/Hg  Baseline 75.4 ± 9.5 76.1 ± 9.6 0.22  Follow-up 76.1 ± 8.6 76.4 ± 7.9 0.24  Change 0.7 ± 9.4 0.3 ± 9.5 0.51 Glucose, mg/dl  Baseline 107.6 ± 26.3 111.3 ± 38.5 0.73  Follow-up 111.9 ± 30.53 114.6 ± 35.9 0.83  Change 4.5 ± 29.5 3.2 ± 34.0 0.95 LDL-C, mg/dl  Baseline 101.4 ± 35.9 105.5 ± 35.4 0.14  Follow-up 83.4 ± 27.4 83.0 ± 27.8 0.70  Change –18.5 ± 38.2 –22.5 ± 38.3 0.13 CRP, mg/l  Baseline 2.3 (1.2, 5.8) 2.3 (1.1, 5.2) 0.29  Follow-up 1.7 (0.8, 4.3) 1.6 (0.7, 3.8) 0.24  Change –0.2 (–2.2, 0.7) –0.4 (–1.9, 0.4) 0.28 . AF . Non-AF . p-value . Triglycerides, mg/dl  Baseline 139.0 (97.3, 211.3) 137.0 (97.8, 193.0) 0.66  Follow-up 131.8 (99.7, 184.0) 127.1 (94.2, 171.8) 0.24  Change –16.0 ± 70.5 –12.4 ± 70.07 0.65 HDL-C, mg/dl  Baseline 42.2 ± 11.5 43.4 ± 11.8 0.24  Follow-up 48.0 ± 15.1 48.2 ± 14.7 0.73  Change 5.8 ± 10.8 4.9 ± 10.2 0.68 Systolic blood pressure, mm/Hg  Baseline 127.0 ± 15.43 128.5 ± 16.24 0.34  Follow-up 130.3 ± 15.1 129.4 ± 13.7 0.24  Change 3.3 ± 16.01 0.9 ± 15.4 0.08 Diastolic blood pressure, mm/Hg  Baseline 75.4 ± 9.5 76.1 ± 9.6 0.22  Follow-up 76.1 ± 8.6 76.4 ± 7.9 0.24  Change 0.7 ± 9.4 0.3 ± 9.5 0.51 Glucose, mg/dl  Baseline 107.6 ± 26.3 111.3 ± 38.5 0.73  Follow-up 111.9 ± 30.53 114.6 ± 35.9 0.83  Change 4.5 ± 29.5 3.2 ± 34.0 0.95 LDL-C, mg/dl  Baseline 101.4 ± 35.9 105.5 ± 35.4 0.14  Follow-up 83.4 ± 27.4 83.0 ± 27.8 0.70  Change –18.5 ± 38.2 –22.5 ± 38.3 0.13 CRP, mg/l  Baseline 2.3 (1.2, 5.8) 2.3 (1.1, 5.2) 0.29  Follow-up 1.7 (0.8, 4.3) 1.6 (0.7, 3.8) 0.24  Change –0.2 (–2.2, 0.7) –0.4 (–1.9, 0.4) 0.28 Results expressed as mean ± SD or median (interquartile range) when not normally distributed (triglyceride, CRP). Least-squares mean ± SEM changes of parameters after controlling for baseline levels. AF: atrial fibrillation; CRP: C-reactive protein; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol. Open in new tab Table 2. Measures of risk factor control in AF and non-AF patients. . AF . Non-AF . p-value . Triglycerides, mg/dl  Baseline 139.0 (97.3, 211.3) 137.0 (97.8, 193.0) 0.66  Follow-up 131.8 (99.7, 184.0) 127.1 (94.2, 171.8) 0.24  Change –16.0 ± 70.5 –12.4 ± 70.07 0.65 HDL-C, mg/dl  Baseline 42.2 ± 11.5 43.4 ± 11.8 0.24  Follow-up 48.0 ± 15.1 48.2 ± 14.7 0.73  Change 5.8 ± 10.8 4.9 ± 10.2 0.68 Systolic blood pressure, mm/Hg  Baseline 127.0 ± 15.43 128.5 ± 16.24 0.34  Follow-up 130.3 ± 15.1 129.4 ± 13.7 0.24  Change 3.3 ± 16.01 0.9 ± 15.4 0.08 Diastolic blood pressure, mm/Hg  Baseline 75.4 ± 9.5 76.1 ± 9.6 0.22  Follow-up 76.1 ± 8.6 76.4 ± 7.9 0.24  Change 0.7 ± 9.4 0.3 ± 9.5 0.51 Glucose, mg/dl  Baseline 107.6 ± 26.3 111.3 ± 38.5 0.73  Follow-up 111.9 ± 30.53 114.6 ± 35.9 0.83  Change 4.5 ± 29.5 3.2 ± 34.0 0.95 LDL-C, mg/dl  Baseline 101.4 ± 35.9 105.5 ± 35.4 0.14  Follow-up 83.4 ± 27.4 83.0 ± 27.8 0.70  Change –18.5 ± 38.2 –22.5 ± 38.3 0.13 CRP, mg/l  Baseline 2.3 (1.2, 5.8) 2.3 (1.1, 5.2) 0.29  Follow-up 1.7 (0.8, 4.3) 1.6 (0.7, 3.8) 0.24  Change –0.2 (–2.2, 0.7) –0.4 (–1.9, 0.4) 0.28 . AF . Non-AF . p-value . Triglycerides, mg/dl  Baseline 139.0 (97.3, 211.3) 137.0 (97.8, 193.0) 0.66  Follow-up 131.8 (99.7, 184.0) 127.1 (94.2, 171.8) 0.24  Change –16.0 ± 70.5 –12.4 ± 70.07 0.65 HDL-C, mg/dl  Baseline 42.2 ± 11.5 43.4 ± 11.8 0.24  Follow-up 48.0 ± 15.1 48.2 ± 14.7 0.73  Change 5.8 ± 10.8 4.9 ± 10.2 0.68 Systolic blood pressure, mm/Hg  Baseline 127.0 ± 15.43 128.5 ± 16.24 0.34  Follow-up 130.3 ± 15.1 129.4 ± 13.7 0.24  Change 3.3 ± 16.01 0.9 ± 15.4 0.08 Diastolic blood pressure, mm/Hg  Baseline 75.4 ± 9.5 76.1 ± 9.6 0.22  Follow-up 76.1 ± 8.6 76.4 ± 7.9 0.24  Change 0.7 ± 9.4 0.3 ± 9.5 0.51 Glucose, mg/dl  Baseline 107.6 ± 26.3 111.3 ± 38.5 0.73  Follow-up 111.9 ± 30.53 114.6 ± 35.9 0.83  Change 4.5 ± 29.5 3.2 ± 34.0 0.95 LDL-C, mg/dl  Baseline 101.4 ± 35.9 105.5 ± 35.4 0.14  Follow-up 83.4 ± 27.4 83.0 ± 27.8 0.70  Change –18.5 ± 38.2 –22.5 ± 38.3 0.13 CRP, mg/l  Baseline 2.3 (1.2, 5.8) 2.3 (1.1, 5.2) 0.29  Follow-up 1.7 (0.8, 4.3) 1.6 (0.7, 3.8) 0.24  Change –0.2 (–2.2, 0.7) –0.4 (–1.9, 0.4) 0.28 Results expressed as mean ± SD or median (interquartile range) when not normally distributed (triglyceride, CRP). Least-squares mean ± SEM changes of parameters after controlling for baseline levels. AF: atrial fibrillation; CRP: C-reactive protein; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol. Open in new tab Measures of atheroma burden and vessel dimensions Measures of atheroma burden and vessel dimensions at baseline and on serial evaluation are summarized in Table 3. Atrial fibrillation patients demonstrated less extensive disease at baseline with lower PAV (36.0 ± 8.9 vs. 38.1 ± 8.9%, p = 0.002) and TAV (185.0 ± 88.7 vs. 188.2 ± 81.8 mm3, p = 0.6), but larger lumen volumes (318.2 ± 116.7 vs. 301.8 ± 109.8 mm3, p = 0.04). Despite the presence of less atherosclerosis, EEM volumes did not differ significantly between groups (503.2 ± 184.1 vs. 490.0 ± 168.9 mm3, p = 0.29). Serial analysis indicated that compared with non-atrial fibrillation-patients, atrial fibrillation-patients demonstrated less PAV progression (–0.05 ± 0.4 vs. +0.25 ±0.4 %, p = 0.002) and greater TAV regression (–1.9 ± 1.7 vs. –4.2 ± 1.7 mm3, p < 0.001). Table 3. Baseline values for intravascular ultrasonographic measures and change in values from baseline with propensity score weighting (IPTW). IVUS parameter . AF vs. non-AF . AF (n = 190) . Non-AF (n = 4776) . Mean difference and 95% CI . p-value . Percent atheroma volume, %  Baseline 36.0 ± 8.9 38.1 ± 8.9 2.01 (0.72, 3.31) 0.002  Change from baseline –0.05 ± 0.36 0.25 ± 0.36 0.30 (0.13, 0.46) <0.001   p-value for test of change = 0 0.90 0.49 Total atheroma volume, mm3  Baseline 185.0 ± 88.7 188.2 ± 81.8 3.16 (–8.74, 15.07) 0.60  Change from baseline –1.85 ± 1.74 –4.18 ± 1.73 –2.32 (–3.38, –1.27) <0.001   p-value for test of change = 0 0.29 0.02 Lumen volume, mm3  Baseline 318.2 ± 116.7 301.8 ± 109.8 –16.41 (–32.37, –0.46) 0.04  Change from baseline –2.71 ± 2.86 –8.17 ± 2.83 –5.46 (–7.46, –3.45) <0.001   p-value for test of change=0 0.34 0.004 EEM volume, mm3  Baseline 503.2 ± 184.1 490.0 ± 168.9 –13.25 (–37.83, 11.33) 0.29  Change from baseline –4.64 ± 3.32 –12.25 ± 3.28 –7.61 (–10.11, –5.11) <0.001   p-value for test of change = 0 0.16 <0.001 IVUS parameter . AF vs. non-AF . AF (n = 190) . Non-AF (n = 4776) . Mean difference and 95% CI . p-value . Percent atheroma volume, %  Baseline 36.0 ± 8.9 38.1 ± 8.9 2.01 (0.72, 3.31) 0.002  Change from baseline –0.05 ± 0.36 0.25 ± 0.36 0.30 (0.13, 0.46) <0.001   p-value for test of change = 0 0.90 0.49 Total atheroma volume, mm3  Baseline 185.0 ± 88.7 188.2 ± 81.8 3.16 (–8.74, 15.07) 0.60  Change from baseline –1.85 ± 1.74 –4.18 ± 1.73 –2.32 (–3.38, –1.27) <0.001   p-value for test of change = 0 0.29 0.02 Lumen volume, mm3  Baseline 318.2 ± 116.7 301.8 ± 109.8 –16.41 (–32.37, –0.46) 0.04  Change from baseline –2.71 ± 2.86 –8.17 ± 2.83 –5.46 (–7.46, –3.45) <0.001   p-value for test of change=0 0.34 0.004 EEM volume, mm3  Baseline 503.2 ± 184.1 490.0 ± 168.9 –13.25 (–37.83, 11.33) 0.29  Change from baseline –4.64 ± 3.32 –12.25 ± 3.28 –7.61 (–10.11, –5.11) <0.001   p-value for test of change = 0 0.16 <0.001 Baseline values are reported as mean ± SD, and change values are reported as least-squares mean ± SE controlling for the baseline counterpart and trial. AF: atrial fibrillation; CI: confidence interval; EEM: external elastic membrane; IPTW: inverse probability of treatment weight; IVUS: intravascular ultrasound Open in new tab Table 3. Baseline values for intravascular ultrasonographic measures and change in values from baseline with propensity score weighting (IPTW). IVUS parameter . AF vs. non-AF . AF (n = 190) . Non-AF (n = 4776) . Mean difference and 95% CI . p-value . Percent atheroma volume, %  Baseline 36.0 ± 8.9 38.1 ± 8.9 2.01 (0.72, 3.31) 0.002  Change from baseline –0.05 ± 0.36 0.25 ± 0.36 0.30 (0.13, 0.46) <0.001   p-value for test of change = 0 0.90 0.49 Total atheroma volume, mm3  Baseline 185.0 ± 88.7 188.2 ± 81.8 3.16 (–8.74, 15.07) 0.60  Change from baseline –1.85 ± 1.74 –4.18 ± 1.73 –2.32 (–3.38, –1.27) <0.001   p-value for test of change = 0 0.29 0.02 Lumen volume, mm3  Baseline 318.2 ± 116.7 301.8 ± 109.8 –16.41 (–32.37, –0.46) 0.04  Change from baseline –2.71 ± 2.86 –8.17 ± 2.83 –5.46 (–7.46, –3.45) <0.001   p-value for test of change=0 0.34 0.004 EEM volume, mm3  Baseline 503.2 ± 184.1 490.0 ± 168.9 –13.25 (–37.83, 11.33) 0.29  Change from baseline –4.64 ± 3.32 –12.25 ± 3.28 –7.61 (–10.11, –5.11) <0.001   p-value for test of change = 0 0.16 <0.001 IVUS parameter . AF vs. non-AF . AF (n = 190) . Non-AF (n = 4776) . Mean difference and 95% CI . p-value . Percent atheroma volume, %  Baseline 36.0 ± 8.9 38.1 ± 8.9 2.01 (0.72, 3.31) 0.002  Change from baseline –0.05 ± 0.36 0.25 ± 0.36 0.30 (0.13, 0.46) <0.001   p-value for test of change = 0 0.90 0.49 Total atheroma volume, mm3  Baseline 185.0 ± 88.7 188.2 ± 81.8 3.16 (–8.74, 15.07) 0.60  Change from baseline –1.85 ± 1.74 –4.18 ± 1.73 –2.32 (–3.38, –1.27) <0.001   p-value for test of change = 0 0.29 0.02 Lumen volume, mm3  Baseline 318.2 ± 116.7 301.8 ± 109.8 –16.41 (–32.37, –0.46) 0.04  Change from baseline –2.71 ± 2.86 –8.17 ± 2.83 –5.46 (–7.46, –3.45) <0.001   p-value for test of change=0 0.34 0.004 EEM volume, mm3  Baseline 503.2 ± 184.1 490.0 ± 168.9 –13.25 (–37.83, 11.33) 0.29  Change from baseline –4.64 ± 3.32 –12.25 ± 3.28 –7.61 (–10.11, –5.11) <0.001   p-value for test of change = 0 0.16 <0.001 Baseline values are reported as mean ± SD, and change values are reported as least-squares mean ± SE controlling for the baseline counterpart and trial. AF: atrial fibrillation; CI: confidence interval; EEM: external elastic membrane; IPTW: inverse probability of treatment weight; IVUS: intravascular ultrasound Open in new tab Factors associated with coronary atheroma progression Table 4 describes factors related with PAV progression in those with or without atrial fibrillation. Presence of atrial fibrillation (p = 0.019), greater baseline PAV (p < 0.001), and use of statins (p < 0.001), aspirin (p < 0.001) and angiotensin-converting enzyme inhibitor/angiotensin receptor blocker (p < 0.001) each associated with less coronary atheroma progression. In contrast, diabetes mellitus (p < 0.001), history of peripheral vascular disease (p < 0.001), increasing age (p < 0.001), higher triglycerides (p = 0.001), and concomitant beta-blocker use (p < 0.001) all independently associated with greater PAV progression. Table 4. Multivariable linear regression model for changes in percent atheroma volume. Covariate . Estimated beta-coefficient (95% CI) . p-value . AF, vs. non-AF –1.68 (–2.94, –0.42) 0.019 Baseline PAV –0.62 (–0.70, –0.53) <0.001 History of diabetics 0.97 (0.70, 1.23) <0.001 History of PVD 0.82 (0.46, 1.17) <0.001 Age 0.48 (0.36, 0.60) <0.001 Concomitant statin use –1.37 (–1.75, –0.98) <0.001 Concomitant aspirin use –1.66 (–2.11, –1.22) <0.001 Concomitant ACE-I/ARB use –1.09 (–1.35, –0.83) <0.001 Concomitant beta blocker use 1.29 (0.93, 1.65) <0.001 Change in triglycerides, log 0.21 (0.11, 0.30) 0.001 Covariate . Estimated beta-coefficient (95% CI) . p-value . AF, vs. non-AF –1.68 (–2.94, –0.42) 0.019 Baseline PAV –0.62 (–0.70, –0.53) <0.001 History of diabetics 0.97 (0.70, 1.23) <0.001 History of PVD 0.82 (0.46, 1.17) <0.001 Age 0.48 (0.36, 0.60) <0.001 Concomitant statin use –1.37 (–1.75, –0.98) <0.001 Concomitant aspirin use –1.66 (–2.11, –1.22) <0.001 Concomitant ACE-I/ARB use –1.09 (–1.35, –0.83) <0.001 Concomitant beta blocker use 1.29 (0.93, 1.65) <0.001 Change in triglycerides, log 0.21 (0.11, 0.30) 0.001 Following inverse probability of treatment weight adjustment BMI was forced into the model. There were also significant interaction effects of AF and age, AF and concomitant statin use, AF and concomitant ACE/ARB use, and AF and concomitant beta blocker use (all have p-value < 0.001). Baseline PAV, age and change in triglycerides are per standard deviation. ACE-I: angiotensin-converting enzyme inhibitor; AF: atrial fibrillation; ARB: angiotensin receptor blocker; BMI: body mass index; CHF: congestive heart failure; CI: confidence interval; PAV: percent atheroma volume; PVD: peripheral vascular disease Open in new tab Table 4. Multivariable linear regression model for changes in percent atheroma volume. Covariate . Estimated beta-coefficient (95% CI) . p-value . AF, vs. non-AF –1.68 (–2.94, –0.42) 0.019 Baseline PAV –0.62 (–0.70, –0.53) <0.001 History of diabetics 0.97 (0.70, 1.23) <0.001 History of PVD 0.82 (0.46, 1.17) <0.001 Age 0.48 (0.36, 0.60) <0.001 Concomitant statin use –1.37 (–1.75, –0.98) <0.001 Concomitant aspirin use –1.66 (–2.11, –1.22) <0.001 Concomitant ACE-I/ARB use –1.09 (–1.35, –0.83) <0.001 Concomitant beta blocker use 1.29 (0.93, 1.65) <0.001 Change in triglycerides, log 0.21 (0.11, 0.30) 0.001 Covariate . Estimated beta-coefficient (95% CI) . p-value . AF, vs. non-AF –1.68 (–2.94, –0.42) 0.019 Baseline PAV –0.62 (–0.70, –0.53) <0.001 History of diabetics 0.97 (0.70, 1.23) <0.001 History of PVD 0.82 (0.46, 1.17) <0.001 Age 0.48 (0.36, 0.60) <0.001 Concomitant statin use –1.37 (–1.75, –0.98) <0.001 Concomitant aspirin use –1.66 (–2.11, –1.22) <0.001 Concomitant ACE-I/ARB use –1.09 (–1.35, –0.83) <0.001 Concomitant beta blocker use 1.29 (0.93, 1.65) <0.001 Change in triglycerides, log 0.21 (0.11, 0.30) 0.001 Following inverse probability of treatment weight adjustment BMI was forced into the model. There were also significant interaction effects of AF and age, AF and concomitant statin use, AF and concomitant ACE/ARB use, and AF and concomitant beta blocker use (all have p-value < 0.001). Baseline PAV, age and change in triglycerides are per standard deviation. ACE-I: angiotensin-converting enzyme inhibitor; AF: atrial fibrillation; ARB: angiotensin receptor blocker; BMI: body mass index; CHF: congestive heart failure; CI: confidence interval; PAV: percent atheroma volume; PVD: peripheral vascular disease Open in new tab Relationship between atrial fibrillation and cardiovascular events In this analysis 96 MACEs (three deaths, 72 MIs, 21 strokes) were recorded. Figure 1 describes the association between the presence of atrial fibrillation, time-to-first MACE and time-to-first MI. The univariate [hazard ratio (95% confidence interval (CI) 1.8 (1.39, 2.38), p < 0.001], PAV adjusted [hazard ratio (95% CI) 2.3 (1.75, 3.05), p < 0.001), and full multivariable (hazard ratio (95% CI) 2.2 (1.66, 2.92), p < 0.001] models all demonstrated a significant relationship between the presence of atrial fibrillation and time-to-first MACE. Similarly, the association between atrial fibrillation and MI remained significant following univariate [hazard ratio (95% CI) 1.7 (1.27, 2.36), p < 0.001], PAV adjusted (hazard ratio (95%CI) 2.3 (1.68, 3.18), p < 0.001), and full multivariable [hazard ratio (95% CI) 2.4 (1.74, 3.35), p < 0.001] adjustments. Kaplan–Meier survival curve analysis (Figure 2(a)) illustrates the cumulative incidence of MACEs in patients with or without atrial fibrillation. Patients with atrial fibrillation had a significantly greater two-year cumulative incidence of MACEs (4.4 vs. 2.0% (log–rank test p = 0.02)) and MI (3.3 vs. 1.5% (log–rank test p = 0.05)) compared with patients without atrial fibrillation. Similar differences were observed when groups were compared with regard to the incidence of MI (Figure 2(b)). Figure 1. Open in new tabDownload slide Relationship between atrial fibrillation, MACE and MI. Forest plot illustrating the independent predictors of atheroma progression in patients with atrial fibrillation on multivariable analysis including all univariate predictors and cardiovascular risk factors. MACE; major adverse cardiovascular event; HR: hazard ratio; CI: confidence interval; PAV: percent atheroma volume; AF: atrial fibrillation; IPTW: inverse probability of treatment weight; ACE: angiotensin-converting enzyme; LDL-C: low-density lipoprotein cholesterol Figure 2. Open in new tabDownload slide (a) Kaplan–Meier survival curve analysis of patients with or without atrial fibrillation at baseline and major adverse cardiovascular event. The cumulative incidence of major adverse cardiovascular event in patients with atrial fibrillation was 4.35%, compared with 2.01% in those without atrial fibrillation (p = 0.017). (b) Kaplan–Meier survival curve analysis of patients with or without atrial fibrillation at baseline and myocardial infarction. The cumulative incidence of myocardial infarction in patients with atrial fibrillation was 3.27%, compared with 1.52% in those without atrial fibrillation (p = 0.05). MACE: major adverse cardiovascular event; AF: atrial fibrillation; MI: myocardial infarction Discussion In this pooled analysis of 4966 patients with established CAD, we observed that a history of atrial fibrillation associated with a greater incidence of MACE and MI. This supports prior reports associating atrial fibrillation with atherosclerotic cardiovascular events. These findings remained valid following adjustment for clinical characteristics and medication use, which are likely to influence cardiovascular outcomes. Surprisingly, atrial fibrillation patients harbored less coronary atherosclerosis at baseline and subsequent disease progression on serial imaging. These findings suggest that factors other than the burden of atherosclerotic plaque underlie the association between atrial fibrillation and incident cardiovascular events. The potential relationship between atrial fibrillation and CAD appears to be complex with evidence that atrial fibrillation can both precede5,6 and follow19,20 the onset of the clinically evident CAD. The current findings that the presence of atrial fibrillation identifies patients at a higher risk of cardiovascular events support prior observations from community cohort studies7-9 and in patients undergoing coronary angiography.8 The specific mechanism underlying this association remains poorly understood. Prior studies employing either computed tomography or invasive coronary angiography have variably reported that atrial fibrillation patients have either greater21,22 or less8,23 disease burden compared with patients without atrial fibrillation. In the present analysis, using IVUS derived measures of atheroma volume, we demonstrated less plaque burden in association with atrial fibrillation. Furthermore, we observed less disease progression in atrial fibrillation patients on serial observation, which persisted after controlling for clinical characteristics. The latter finding is particularly striking given prior observations that incident cardiovascular events associate with greater plaque progression.24 The current findings suggest that the association between atrial fibrillation and incident cardiovascular events cannot simply be attributed to the presence of more extensive coronary atherosclerosis. This raises further questions regarding potential mechanistic links promoting adverse cardiovascular outcomes in these patients. A number of potential mechanisms may contribute to these events. There is increasing recognition that inflammation promotes both atrial fibrillation and atherosclerotic events.25,26 While C-reactive protein levels did not differ between the groups in this analysis, and that adjusting for C-reactive protein failed to attenuate the association between atrial fibrillation and MACE, it is possible that underlying plaque composition may have differed between atrial fibrillation and non-atrial fibrillation patients. In fact, increasing recognition of the potential role of inflammation in atrial fibrillation may associate with coronary atherosclerosis that is more inflamed in nature, resulting in a greater risk of adverse cardiovascular events. The association between atrial fibrillation and the development of both a prothrombotic milieu27-29 and potential coronary embolism30 may further confer an enhanced risk of MI. The potential impact of atrial fibrillation related tachyarrhythmias in promoting oxidative stress, microvascular dysfunction and demand induced ischemia also cannot be underestimated.31-33 A number of caveats should be noted with regard to the current analysis. This is a retrospective pooled post hoc analysis of nine clinical trials that used single vessel IVUS examination, prospectively powered to detect changes in plaque volume, but not MACEs. However, MACEs were prospectively and centrally adjudicated. While there was some heterogeneity across studies, all trials were conducted by the same group, using identical inclusion/exclusion criteria, imaging protocols and analysis within the same core laboratory. Although our total sample size was quite large, the number of patients with atrial fibrillation was relatively low. However, since physician diagnosis was used to identify atrial fibrillation patients, it is possible that some patients in the non-atrial fibrillation group might have had paroxysmal atrial fibrillation. In addition, to address this limitation we employed propensity scores, and corresponding inverse probability of treatment weight was calculated through a logistic regression model with baseline atrial fibrillation status as the outcome and using all the related baseline characteristics and medications as covariates. It should be also noted that all patients had presented for clinically indicated coronary angiography, accordingly it is unknown whether the results are applicable to asymptomatic patients with atrial fibrillation. In summary, to our knowledge this is the first report that integrates serial measures of coronary atheroma burden and cardiovascular outcomes in patients with and without atrial fibrillation. The presence of atrial fibrillation independently increased with a greater risk of MI, despite smaller plaque burden and less progression of coronary atheroma. These findings suggest that factors beyond plaque burden are likely to underscore the relationship between atrial fibrillation and CAD. Acknowledgement The authors are grateful for the technical expertise of the Intravascular Core Laboratory of the Cleveland Clinic. Author contribution Study concept and design: OB, RP, EMT, SK, SN, SJN. Acquisition of data: EMT, SN, PS, SJN, KW. Analysis and Interpretation of data: OB, RP, MS, EMT, KW, SK, PS, SN, SJN. Drafting of the manuscript: OB, RP, EMT, SJN. Critically revised the manuscript: SK, PS, SN, MS, PS. Statistical analysis: OB, RP, MS, KW. All gave final approval and agree to be accountable for all aspects of work ensuring integrity and accuracy. 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: SJN reports receiving honoraria from Pfizer, AstraZeneca, Takeda and Merck Schering-Plough, consultancy fees from AstraZeneca, Pfizer, Roche, Novo-Nordisk, Merck Schering-Plough, Liposcience, and Anthera Pharmaceuticals, and research support from Astra Zeneca, Lipid Sciences, Novartis, and Resverlogix. SEN has received research support from AstraZeneca, Eli Lilly, Pfizer, Takeda, Sankyo, and Sanofi-Aventis. He has consulted for a number of pharmaceutical companies without financial compensation. All honoraria, consulting fees or any other payments from any for-profit entity are paid directly to charity, so that neither income nor any tax deduction is received. All other authors have no disclosures. Funding The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Internal departmental funding for this analysis was provided. REVERSAL, CAMELOT, ILLUSTRATE funded by Pfizer. ACTIVATE funded by Sankyo. ASTEROID and SATURN funded by Astra-Zeneca. AQUARIUS funded by Novartis Pharmaceuticals. PERISCOPE funded by Takeda. STRADIVARIUS funded by Sanofi-Aventis. None of these pharmaceutical companies sponsored the current analysis. References 1 Hobbs FR , Taylor CJ, Jan Geersing Get al. European Primary Care Cardiovascular Society (EPCCS) consensus guidance on stroke prevention in atrial fibrillation (SPAF) in primary care . Eur J Prev Cardiol 2016 ; 23 : 460 – 473 . Google Scholar Crossref Search ADS PubMed WorldCat 2 January CT, Wann LS, Alpert JS, et al. ACC/AHA Task Force Members. 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: Executive summary: A report of the American College of Cardiology/American Heart Association Task Force on practice guidelines and the Heart Rhythm Society. Circulation 2014; 130: e199–e267 . 3 Pallisgaard JL , Schjerning AM, Lindhardt TBet al. Risk of atrial fibrillation in diabetes mellitus: A nationwide cohort study . Eur J Prev Cardiol 2016 ; 23 : 621 – 627 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Kannel WB, Wolf PA, Benjamin EJ, et al. Prevalence, incidence, prognosis, and predisposing conditions for atrial fibrillation: Population-based estimates. Am J Cardiol 1998; 82: 2N–9N . 5 Benjamin EJ , Levy D, Vaziri SMet al. Independent risk factors for atrial fibrillation in a population-based cohort. The Framingham Heart Study . JAMA 1994 ; 271 : 840 – 844 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Krahn AD , Manfred J, Tate RBet al. The natural history of atrial fibrillation: Incidence, risk factors, and prognosis in the Manitoba Follow-Up Study . Am J Med 1995 ; 98 : 476 – 484 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Miyasaka Y , Barnes ME, Gersh BJet al. Coronary ischemic events after first atrial fibrillation: Risk and survival . Am J Med 2007 ; 120 : 357 – 363 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Marte T , Saely CH, Schmid Fet al. Effectiveness of atrial fibrillation as an independent predictor of death and coronary events in patients having coronary angiography . Am J Cardiol 2009 . 103: 36–40 . Google Scholar OpenURL Placeholder Text WorldCat 9 Soliman EZ , Safford MM, Muntner Pet al. Atrial fibrillation and the risk of myocardial infarction . JAMA Intern Med 2014 ; 174 : 107 – 114 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Nissen SE , Tuzcu EM, Schoenhagen Pet al. REVERSAL Investigators . Effect of intensive compared with moderate lipid-lowering therapy on progression of coronary atherosclerosis: A randomized controlled trial . JAMA 2004 ; 291 : 1071 – 1080 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Nissen SE , Nicholls SJ, Sipahi Iet al. ASTEROID Investigators . Effect of very high-intensity statin therapy on regression of coronary atherosclerosis: The ASTEROID trial . JAMA 2006 ; 295 : 1556 – 1565 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Nicholls SJ , Ballantyne CM, Barter PJet al. Effect of two intensive statin regimens on progression of coronary disease . N Engl J Med 2011 ; 365 : 2078 – 2087 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Nicholls SJ , Bakris GL, Kastelein JJet al. Effect of aliskiren on progression of coronary disease in patients with prehypertension: The AQUARIUS randomized clinical trial . JAMA 2013 ; 310 : 1135 – 1144 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Nissen SE , Tuzcu EM, Libby Pet al. CAMELOT Investigators . Effect of antihypertensive agents on cardiovascular events in patients with coronary disease and normal blood pressure: The CAMELOT study: A randomized controlled trial . JAMA 2004 ; 292 : 2217 – 2225 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Nissen SE , Tuzcu EM, Brewer HBet al. ACTIVATE Investigators . Effect of ACAT inhibition on the progression of coronary atherosclerosis . N Engl J Med 2006 ; 354 : 1253 – 1263 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Nissen SE , Tardif JC, Nicholls SJet al. ILLUSTRATE Investigators . Effect of torcetrapib on the progression of coronary atherosclerosis . N Engl J Med 2007 ; 356 : 1304 – 1316 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Nissen SE , Nicholls SJ, Wolski Ket al. STRADIVARIUS Investigators . Effect of rimonabant on progression of atherosclerosis in patients with abdominal obesity and coronary artery disease: The STRADIVARIUS randomized controlled trial . JAMA 2008 ; 299 : 1547 – 1560 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Nissen SE , Nicholls SJ, Wolski Ket al. PERISCOPE Investigators . Comparison of pioglitazone vs. glimepiride on progression of coronary atherosclerosis in patients with type 2 diabetes: The PERISCOPE randomized controlled trial . JAMA 2008 ; 299 : 1561 – 1573 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Crenshaw BS , Ward SR, Granger CBet al. Atrial fibrillation in the setting of acute myocardial infarction: The GUSTO-I experience. Global Utilization of Streptokinase and TPA for Occluded Coronary Arteries . J Am Coll Cardiol 1997 ; 30 : 406 – 413 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Schmitt J , Duray G, Gersh BJet al. Atrial fibrillation in acute myocardial infarction: A systematic review of the incidence, clinical features and prognostic implications . Eur Heart J 2009 ; 30 : 1038 – 1045 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Nucifora G , Schuijf JD, Tops LFet al. Prevalence of coronary artery disease assessed by multislice computed tomography coronary angiography in patients with paroxysmal or persistent atrial fibrillation . Circ Cardiovasc Imaging 2009 ; 2 : 100 – 106 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Weijs B , Pisters R, Haest RJet al. Patients originally diagnosed with idiopathic atrial fibrillation more often suffer from insidious coronary artery disease compared to healthy sinus rhythm controls . Heart Rhythm 2012 ; 9 : 1923 – 1929 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Cameron A , Schwartz MJ, Kronmal RAet al. Prevalence and significance of atrial fibrillation in coronary artery disease (CASS Registry) . Am J Cardiol 1988 ; 61 : 714 – 717 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Nicholls SJ , Hsu A, Wolski Ket al. Intravascular ultrasound-derived measures of coronary atherosclerotic plaque burden and clinical outcome . J Am Coll Cardiol 2010 ; 55 : 2399 – 2407 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Chung MK , Martin DO, Sprecher Det al. C-reactive protein elevation in patients with atrial arrhythmias: Inflammatory mechanisms and persistence of atrial fibrillation . Circulation 2001 ; 104 : 2886 – 2891 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Ridker PM , Rifai N, Stampfer MJet al. Plasma concentration of interleukin-6 and the risk of future myocardial infarction among apparently healthy men . Circulation 2000 ; 101 : 1767 – 1772 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Lim HS , Willoughby SR, Schultz Cet al. Effect of atrial fibrillation on atrial thrombogenesis in humans: Impact of rate and rhythm . J Am Coll Cardiol 2013 ; 61 : 852 – 860 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Willoughby SR , Roberts-Thomson RL, Lim HSet al. Atrial platelet reactivity in patients with atrial fibrillation . Heart Rhythm 2010 ; 7 : 1178 – 1183 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Lim HS , Willoughby SR, Schultz Cet al. Successful catheter ablation decreases platelet activation and improves endothelial function in patients with atrial fibrillation . Heart Rhythm 2014 ; 11 : 1912 – 1918 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Shibata T , Kawakami K, Noguchi Tet al. Prevalence, clinical features, and prognosis of acute myocardial infarction due to coronary artery embolism . Circulation 2015 ; 132 : 241 – 250 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Goette A , Bukowska A, Dobrev Det al. Acute atrial tachyarrhythmia induces angiotensin II type 1 receptor-mediated oxidative stress and microvascular flow abnormalities in the ventricles . Eur Heart J 2009 ; 30 : 1411 – 1420 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Saaby L , Poulsen TS, Hosbond Set al. Classification of myocardial infarction: Frequency and features of type 2 myocardial infarction . Am J Med 2013 ; 126 : 789 – 797 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Saaby L , Poulsen TS, Diederichsen ACet al. Mortality rate in type 2 myocardial infarction: Observations from an unselected hospital cohort . Am J Med 2014 ; 127 : 295 – 302 . Google Scholar Crossref Search ADS PubMed WorldCat © The European Society of Cardiology 2017 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 2017