International Journal of Obesity (2015), 1–9 © 2015 Macmillan Publishers Limited All rights reserved 0307-0565/15 www.nature.com/ijo

ORIGINAL ARTICLE

Changes in body weight and pulse: outcome events in overweight and obese subjects with cardiovascular disease in the SCOUT trial RV Seimon1, D Espinoza2, N Finer3, WPT James4, UF Legler5, W Coutinho6, AM Sharma7, L Van Gaal8, AP Maggioni9, A Sweeting1, C Torp-Pedersen10, V Gebski2 and ID Caterson1 BACKGROUND/OBJECTIVES: The Sibutramine Cardiovascular OUTcomes (SCOUT) trial showed a significantly increased relative risk of nonfatal cardiovascular events, but not mortality, in overweight and obese subjects receiving long-term sibutramine treatment with diet and exercise. We examined the relationship between early changes (both increases and decreases) in pulse rate, and the impact of these changes on subsequent cardiovascular outcome events in both the placebo and sibutramine groups. SUBJECTS/METHODS: 9804 males and females, aged ⩾ 55 years, with a body mass index of 27–45 kg m−2 were included in this current subanalysis of the SCOUT trial. Subjects were required to have a history of cardiovascular disease and/or type 2 diabetes mellitus with at least one cardiovascular risk factor, to assess cardiovascular outcomes. The primary outcome event (POE) was a composite of nonfatal myocardial infarction, nonfatal stroke, resuscitated cardiac arrest or cardiovascular death. Time-to-event analyses of the POE were performed using Cox regression models. RESULTS: During the initial 6-week sibutramine treatment period, the induced pulse rate increase was related to weight change (1.9 ± 7.7 beats per minute (bpm) with weight increase; 1.4 ± 7.3 bpm, 0–5 kg weight loss; 0.6 ± 7.4 bpm, ⩾ 5 kg weight loss). Throughout the subsequent treatment period, those continuing on sibutramine showed a consistently higher mean pulse rate than the placebo group. There was no difference in POE rates with either an increase or decrease in pulse rate over the lead-in period, or during lead-in baseline to 12 months post randomization. There was also no relationship between pulse rate at lead-in baseline and subsequent cardiovascular events in subjects with or without a cardiac arrhythmia. CONCLUSION: Baseline pulse rate and changes in pulse rate may not be an important modifier nor a clinically useful predictor of outcome in an individual elderly cardiovascular obese subject exposed to weight management. International Journal of Obesity advance online publication, 3 February 2015; doi:10.1038/ijo.2014.211

INTRODUCTION Heart rate (pulse rate) has become an accepted modifiable risk factor for cardiovascular disease (CVD).1 Previous studies have demonstrated a relationship between elevated resting heart rate and the risk for CVD in the overall population,2 and in patients with coronary artery disease with or without comorbidities, including hypertension and/or type 2 diabetes mellitus.3–5 This relationship is strong, graded and independent of other cardiovascular risk factors. However, the threshold above which risk increases and the relationship between alterations in heart rate and CVD outcome are still disputed.4 Analysis of the BEAUTIFUL (morBidity-mortality EvAlUaTion of the If inhibitor ivabradine in patients with coronary disease and left ventricULar dysfunction) and INVEST (International Verapamil-Trandolapril Study) study, identified ⩾ 70 beats per minute (bpm)5 and ⩾ 75 bpm,4 respectively, as a threshold for adverse cardiovascular outcomes. However, the unexpected and controversial results of SIGNIFY6

and SHIFT7,8 have caused doubt that heart rate control is always beneficial in a stable cardiovascular population. Little is known regarding this risk relationship in a contemporary elderly, obese population with several comorbidities, and an activated sympathetic nervous system displayed as elevated resting heart rate within the framework of recent guidelines both for the primary and secondary prevention of CVD.9–11 Sibutramine hydrochloride monohydrate (sibutramine) is a norepinephrine and serotonin reuptake inhibitor that induces satiety, resulting in reduced food intake and increased energy expenditure.12,13 In some individuals, sibutramine also increases blood pressure and/or pulse, in part, due to its sympathomimetic effects.14,15 The Sibutramine Cardiovascular OUTcomes (SCOUT) trial assessed the morbidity and mortality benefits of weight management in over 10 000 overweight and obese subjects with and without established CVD, hypertension and type 2 diabetes mellitus, all of whom were already on medication for their

1 The Boden Institute of Obesity, Nutrition Exercise & Eating Disorders, University of Sydney, Sydney, New South Wales, Australia; 2National Health and Medical Research Council Clinical Trials Centre, University of Sydney, Sydney, New South Wales, Australia; 3National Centre for Cardiovascular Prevention and Outcomes, University College London, Institute for Cardiovascular Science, London, UK; 4London School of Hygiene and Tropical Medicine, London, UK; 5Special Vocational College for Handicapped Persons, Mainz, Germany; 6Catholic University of Rio de Janeiro and State Institute of Diabetes and Endocrinology, Rio de Janeiro, Brazil; 7Department of Medicine, University of Alberta, Edmonton, Alberta, Canada; 8Department of Diabetology, Metabolism and Clinical Nutrition, Antwerp University Hospital, Antwerp, Belgium; 9Associazione Nazionale Medici Cardiologi Ospedalieri Research Center, Florence, Italy and 10Department of Cardiology, Gentofte University Hospital, Hellerup, Denmark. Correspondence: Dr R Seimon, The Boden Institute of Obesity, Nutrition Exercise & Eating Disorders, The University Of Sydney, 92-94 Parramatta Road, Exercise & Eating Disorders, Camperdown, New South Wales 2050, Australia. E-mail: [email protected] Received 29 August 2014; revised 25 November 2014; accepted 5 December 2014; accepted article preview online 18 December 2014

Sibutramine, weight loss, pulse rate and cardiovascular outcomes RV Seimon et al

2 comorbidities, which could be adjusted once the run-in period had been completed.16 The primary results of SCOUT reported a 16% relative risk increase in primary outcome events for long-term sibutramine therapy and denying therefore the study hypothesis; this analysis, however, included the full intention-to-treat (ITT) population and did not take into account the variable responses in body weight and pulse rate seen in this overweight and obese population.16 We recently demonstrated a complex interaction between the effects of sibutramine on both blood pressure and weight loss and subsequent cardiovascular events: modest weight loss and lower blood pressure each reduced the incidence of cardiovascular events but the combination of early marked weight loss and rapid blood pressure reduction seemed to be particularly harmful in an obese elderly population with CVDs.17 The aim of the retrospective analysis presented here is to characterize and compare the relationship between early changes (both increases and decreases) in body weight and pulse rate, and their impact on subsequent cardiovascular outcome events in the SCOUT population. PATIENTS AND METHODS The design and the primary results of SCOUT, a multinational, prospective, randomized, double-blind study conducted according to the principles of the International Conference on Harmonization and the Declaration of Helsinki, have been described in detail previously.16,18 Locally appointed ethics committees approved the protocol, and informed consent was obtained from all subjects. The trial included males and females aged 55 years and older, with a body mass index of 27–45 kg m − 2. Subjects were required to have a history of a clinically stable CVD (defined as coronary artery disease, stroke or peripheral arterial occlusive disease) and/or type 2 diabetes mellitus with at least one other cardiovascular risk factor (hypertension, dyslipidemia, current smoking or diabetic nephropathy).16 As described previously, all the subjects underwent an initial 6-week single-blind lead-in period during which they received 10 mg sibutramine per day as well as advice regarding diet and exercise, in addition to their usual pharmacotherapy, (dosage of which also remained unchanged) so that subjects with marked increases in blood pressure or pulse rate due to sibutramine could be identified and excluded from the prolonged post-randomization period of observation.16 Eligible subjects were then randomized, in a double-blind manner, to receive 10–15 mg of sibutramine or a matching placebo per day. All subjects continued to participate in individualized diet and exercise programs designed to result in an energy deficit of 600 kcal per day.16 Heart rate was measured after the subject had been sitting, resting for at least 5 min. Sitting pulse rate was measured twice at least 1 min apart by palpation of the radial or brachial artery for 1 min. The mean value was used for assessment. All changes in vital signs during the 6-week lead-in sibutramine treatment period were attributed to study interventions as pre-existing pharmacotherapy for comorbidities remained unchanged. The primary outcome event (POE) was the time from randomization, to the first occurrence of nonfatal myocardial infarction, nonfatal stroke, resuscitation after cardiac arrest or cardiovascular death. Standard-of-care non-pharmacological and pharmacological treatment recommendations based on the respective national and international guidelines for the management of comorbidities and for primary and secondary prevention of CVD were applied throughout the randomized period.1,9,11,19–22 From the overall cohort of 10 744 subjects, 940 subjects either withdrew from the 6-week sibutramine lead-in period or, according to protocol, were not randomized by the investigators because of marked increases in blood pressure or pulse rate during this initial treatment with sibutramine. Out of the remaining 9804 subjects, 1272 had a cardiac arrhythmia (for example, atrial fibrillation/flutter) or a pacemaker or implantable cardioverterdefibrillator (Table 1).

Statistical analysis Demographic and baseline characteristics at lead-in period baseline were summarized for those who on randomization constituted the ITT population, by treatment group (randomized sibutramine, randomized placebo). International Journal of Obesity (2015) 1 – 9

Table 1. Demographics and clinical characteristics of the subjects at the randomized phase baseline Variable

Placebo

Sibutramine

Gender, % male Age (years) Weight (kg) BMI (kg m − 2) SBP (mm Hg) DBP (mm Hg) Pulse (bpm)

2843/4898 (58%) 63.3 ± 6.1 96.2 ± 15.5 (4897) 34.4 ± 4.5 (4897) 138.2 ± 12.6 (4897) 77.9 ± 8.4 (4897) 71.1 ± 10.1 (4897)

2807/4906 (57.2%) 63.2 ± 6.1 96.3 ± 15.4 (4905) 34.5 ± 4.6 (4905) 138.2 ± 12.9 (4905) 77.8 ± 8.4 (4905) 71.1 ± 10.2 (4904)

Medical history Overall cardiovascular disease Revascularization procedure Type 2 diabetes mellitus Cardiac arrhythmia

3977/4884 1828/4898 4139/4884 662/4898

(81.4%) (37.3%) (84.7%) (13.5%)

3971/4889 1786/4906 4167/4889 610/4906

(81.2%) (36.4%) (85.2%) (12.4%)

Concomitant medication Angiotensin receptor blockers β-blockers Calcium channel blocker Diuretics Aspirin Statins

3792/4898 3034/4898 1796/4898 2360/4898 3877/4898 3235/4898

(77.4%) (61.9%) (36.7%) (48.2%) (79.2%) (66.0%)

3826/4906 2983/4906 1861/4906 2310/4906 3832/4906 3288/4906

(78.0%) (60.8%) (37.9%) (47.1%) (78.1%) (67.0%)

Abbreviations: BMI, body mass index; bpm, beats per minute; DBP, diastolic blood pressure; SBP, systolic blood pressure. ± values are mean ± s.d. All subjects received sibutramine during the 6-week lead-in period.

Changes in weight and pulse rate for individual subjects were computed from the lead-in baseline (week − 6) to randomization (that is, end of 6-week sibutramine lead-in period) and to 12 months post randomization. Data were summarized within each category of interest as mean and s.d. for continuous variables and as the number of subjects and percentage (%) for categorical variables. Mean changes in vital sign measurements from lead-in baseline to randomization and to 12 months post randomization were evaluated using paired t-tests. Time-to-event analyses were conducted looking at the effect of pulse change (no change (±1 bpm), increase or decrease) on POE rates for the ITT population using Cox models, with factors for treatment (that is, 6-week sibutramine treatment followed by either placebo (lifestyle+sibutramine followed by placebo; placebo group) or continued sibutramine (lifestyle+ sibutramine followed by sibutramine; long-term sibutramine group)), country, sex and age as covariates. Pulse change (no change (±1 bpm), increase or decrease) for individual subjects calculated over (a) the lead-in baseline to randomization and (b) the lead-in baseline to 12 months post randomization in subjects with and without cardiac arrhythmia was assessed. In each case the reference group, when comparing categories, was the no pulse change (±1 bpm) group. Estimates of hazard ratios (HR), 95% confidence intervals (CI) and log-rank P-values were calculated within the Cox model framework. Analyses were conducted for (a) all POEs reported during the randomization to a 5-year period and (b) all POEs that occurred after 12 months post randomization (subjects who had a POE or were censored in the first 12 months were not included in this analysis). Further stratified analysis looking at the effect of treatment within pulse change categories was conducted for pulse change categorized by 2.5 bpm change increments for individual subjects over (a) the lead-in baseline to randomization and (b) the lead-in baseline to 12 months post randomization in subjects with and without cardiac arrhythmia. All statistical analyses were performed using SAS, version 9.2 (SAS Institute, Cary, NC, USA). All statistical tests were conducted at the (two tail) 0.05 level of significance. Body weight and vital sign changes were imputed using a last observation carried forward approach if measurements were not recorded at the indicated time points after lead-in baseline.

RESULTS The demographics and clinical characteristics at the 6-week leadin baseline for the overall study cohort (N = 9804) and the two randomized treatment groups (6-week sibutramine followed by placebo (N = 4898); 6-week sibutramine followed by sibutramine © 2015 Macmillan Publishers Limited

Sibutramine, weight loss, pulse rate and cardiovascular outcomes RV Seimon et al

(N = 4906)) have been described previously14,18 and are summarized in Table 1. Two thousand one hundred and twenty-two out of 4898 in the placebo and 2034/4906 in the sibutramine group completed the whole treatment period. The most common primary reason for subjects prematurely discontinuing the study therapy during the treatment period was withdrawal of consent (which in practice is usually related to the individual's discontent with the amount of weight lost). Treatment groups were well balanced with the exception that subjects in the long-term sibutramine treatment arm were more likely already to have peripheral artery occlusive disease. At lead-in baseline, mean body weight was 96 kg, body mass index was 34.5 kg m − 2, mean systolic blood pressure was 138 mm Hg and diastolic blood pressure 78 mm Hg and pulse rate was 71 bpm, in both the subsequently randomized treatment groups.

Changes in weight and pulse rate during the entire trial period As described previously,17 during the 6-week lead-in period, from lead-in baseline (week-6) to randomization, when all subjects received sibutramine and comorbidity related pharmacotherapy remained unaltered, there was a mean weight loss of 2.5 kg. Following randomization, there was further weight loss in the sibutramine group (maximum mean additional weight loss 1.8 kg at 12 months) and a small average increase in weight in the placebo group (0.6 kg at 12 months). After the 12-month period, both the groups showed only a small increase in mean weight. During the 6-week lead-in period, the mean pulse rate increased by 1.4 bpm (Figure 1). Following randomization, mean pulse rate increased further in the sibutramine group, but decreased in those randomized to placebo. Throughout the entire treatment period, those who continued on sibutramine showed a consistently higher mean pulse rate than the placebo group; mean differences between the groups ranged from 2.2–3.7 bpm (Figure 1).

Figure 1. The average pulse rate at each visit from the time of the lead-in period (Wk-6) to the final visit (60 months) for all the randomized subjects. All subjects underwent an initial 6-week (Wk-6 to randomization), single-blind lead-in period during which they received 10 mg of sibutramine per day; subjects considered eligible were then randomized to sibutramine or placebo. Analyses were performed on the data obtained from the ITT population using t-test. Pulse rate, P = 0.05 at months 30, 36 and 39, P = 0.01 at months 12, 18–24, 45 and 54, Po 0.0001 at months 1–9, 15, 27 and 33 for the comparison between sibutramine and placebo. Wk, week; R, randomization; M, month. © 2015 Macmillan Publishers Limited

3 Interactions between pulse rate and weight The relationship between categorical weight change (subjects who gained weight, lost 0 to o5 kg, lost 5 kg or more) during the sibutramine treatment lead-in period and subsequent pulse rate change from randomization throughout the trial is shown for both the randomized ITT (Figure 2a) and completer (Figure 2b) populations. Similarly, categorical weight change during lead-in baseline to 12 months and subsequent pulse rate change from randomization is shown for the randomized ITT (Figure 2c) and completer (Figure 2d) populations. Regardless of weight change, in all analyses, those on sibutramine showed a significantly higher pulse rate when compared with those on placebo, throughout the trial. In both the ITT and the completer populations, those subsequently maintained on sibutramine who lost weight during lead-in baseline to randomization period (Figures 2a and b), showed a graded and smaller increase in pulse rate with greater weight loss over the first 12 months. This was also observed in those who lost weight from lead-in baseline to 12 months post randomization (Figures 2c and d); they also showed a graded and smaller increase in pulse rate with greater weight loss over the 24 months. However, in the placebo group there was almost no effect of weight loss on pulse rate. In contrast, in both the placebo and sibutramine groups who gained weight during lead-in baseline to randomization, there was a consistent subsequent fall in the pulse rate. Relationship between pulse rate at lead-in baseline (−6 weeks) and incidence of POE Subjects without a cardiac arrhythmia showed no relationship between pulse rate at lead-in baseline and subsequent cardiovascular events (Figure 3a). In those with a pre-existing cardiac arrhythmia, the relationship was much less consistent; patients on long-term sibutramine who had had a pulse rate of 465 to ⩽ 70 bpm at lead-in baseline showed an increased risk of POE (HR: 2.17 CI: 1.18–3.99, P = 0.01), when compared with placebo, but such an effect could not be seen in subjects within any other lead-in baseline pulse category (Figure 3b). However, in neither case, was there an interaction effect between the different lead-in baseline pulse change categories and sibutramine (subjects with cardiac arrhythmia: P = 0.59 and without cardiac arrhythmia P = 0.47). When adjusted for treatment, country, age and sex, in subjects without cardiac arrhythmia there was an increased risk of POE with a pulse rate of 470 bpm (HR: 1.18 CI: 1.03–1.35, P = 0.02), 475 (HR: 1.19 CI: 1.03–1.38, P = 0.02) and 485 (HR: 1.26 CI: 1.01– 1.58, P = 0.04) at lead-in baseline, when each of those with a higher pulse rate value (⩽70, ⩽ 75 and ⩽ 85 bpm) was compared with all those below this pulse rate. For subjects with a preexisting cardiac arrhythmia, pulse rate of 470 bpm (HR: 1.42 CI: 1.08–1.86, P = 0.01) and 475 (HR: 1.44 CI: 1.09–1.90, P = 0.01) at lead-in baseline, also showed an increased risk of POE when compared with a lead-in baseline pulse rate of ⩽ 70 and ⩽ 75 bpm, respectively. Detailed analysis showing participants in different pulse rate categories at lead-in baseline and changes in pulse rate categories during the lead-in period by various comorbidities (age, β-blockade, baseline cardiac arrhythmias, gender and treatment) are presented in Supplementary Tables S1 and S2. Relationship between changes in pulse rate and the incidence of POE There were no differences in POE rates in subjects who either increased or decreased pulse rate over the lead-in period, or during lead-in baseline to 12 months post randomization (Table 2). There was also no difference in various POE components by pulse rate categories at lead-in baseline or 12 months International Journal of Obesity (2015) 1 – 9

Sibutramine, weight loss, pulse rate and cardiovascular outcomes RV Seimon et al

4

Figure 2. Mean change in pulse rate (bpm) from the randomization of patients into the sibutramine and placebo groups but categorized by the extent of weight change from lead-in period to randomization (a and b) and from lead-in period to 12 months post randomization (c and d), for ITT and completer population, in subjects who gained weight (40 kg), lost 0 to o5 kg, lost 5 kg or more (⩾ −5 kg). R, randomization; M, month.

Figure 3. Subgroup analysis looking at the effect of treatment on primary outcome events stratified by pulse rate categories at lead-in baseline (−6 weeks) in subjects without (a) and with (b) cardiac arrhythmias. Estimates of hazard ratios (95% CI) and log-rank P-values were calculated using the Cox model, adjusted for various factors. Lead-in baseline pulse categories: ⩽ 65, 465– ⩽ 70, 470– ⩽ 75, 475– ⩽ 80, 480– ⩽ 85 and 485 bpm. HR, hazard ratios; CI, confidence intervals.

International Journal of Obesity (2015) 1 – 9

© 2015 Macmillan Publishers Limited

Sibutramine, weight loss, pulse rate and cardiovascular outcomes RV Seimon et al

5 Table 2. Primary outcome events by absolute change categories in pulse rate responsiveness during lead-in period (−6 weeks) to randomization and over subsequent 12 months (note: subjects with cardiac arrhythmias were excluded from this analysis) P-value

Model 2a HR (95%CI)

P-value

Model 3a HR (95%CI)

P-value

Absolute change in pulse rate during lead-in to randomization change Decrease 2956 141 (9.63%) 160 (10.72%) 301 (10.18%) 1.10 (0.88–1.38) No change (±1) 1060 51 (9.51%) 47 (8.97%) 98 (9.25%) 1.00 Increase 4513 196 (8.77%) 233 (10.23%) 429 (9.51%) 1.01 (0.81–1.26)

0.40 0.26b 0.93

1.07 (0.85–1.34) 1.00 1.00 (0.80–1.25)

0.57 0.38b 0.99

1.10 (0.88–1.38) 1.00 1.01 (0.81–1.26)

0.41 0.26b 0.94

Absolute change in pulse rate from lead-in baseline to 12 months post randomization Decrease 3191 128 (6.83%) 116 (8.81%) 244 (7.65%) 0.93 (0.70–1.23) No change (±1) 721 32 (8.40%) 28 (8.24%) 60 (8.32%) 1.00 Increase 3251 86 (6.86%) 153 (7.66%) 239 (7.35%) 0.88 (0.66–1.17)

0.59 0.57b 0.37

0.94 (0.71–1.25) 1.00 0.88 (0.66–1.17)

0.68 0.44b 0.37

0.94 (0.70–1.24) 1.00 0.86 (0.65–1.14)

0.64 0.36b 0.29

Pulse categories (bpm)

N

Placebo

Event rate (%) Sibutramine

Total

Model 1a HR (95% CI)

Abbreviations: CI, confidence interval; HR, hazard ratio. Estimates of hazard ratios (95% CI) and log-rank P-values were calculated using the Cox model. aModel 1: not adjusted for any factors; Model 2: adjusted for sibutramine or placebo treatment, country, sex and age; Model 3: adjusted for treatment only. Pulse categories: no change (±1 bpm), increase or decrease. Data are absolute number (%). Reference group: no change (±1 bpm). P-values refer to the HR of each pulse category compared with the reference group. bIndicated P-value for trend.

Table 3A.

Primary outcome event components by pulse rate categories at lead-in period (−6 weeks) and at 12 months ⩽ 65

70–75

75–80

80–85

485

P-valuea

1680 (19.2%) 195 (18.6%)

1203 (13.7%) 151 (14.4%)

813 (9.3%) 107 (10.2%)

799 (9.1%) 118 (11.2%)

0.19

1786 (19.1%) 89 (19.7%)

1286 (13.8%) 68 (15.0%)

874 (9.3%) 46 (10.2%)

867 (9.3%) 50 (11.1%)

0.39

1806 (19.1%) 69 (19.2%)

1305 (13.8%) 49 (13.6%)

891 (9.4%) 29 (8.1%)

879 (9.3%) 38 (10.6%)

0.93

1873 (19.1%) 2 (11.1%)

1350 (13.8%) 4 (22.2%)

917 (9.4%) 3 (16.7%)

914 (9.3%) 3 (16.7%)

0.39

1840 (19.2%) 35 (15.8%)

1324 (13.8%) 30 (13.5%)

891 (9.3%) 29 (13.1%)

890 (9.3%) 27 (12.2%)

0.19

1355 (18.0%) 132 (18.8%)

1356 (18.0%) 109 (15.5%)

1041 (13.8%) 84 (12.0%)

670 (8.9%) 70 (10.0%)

818 (10.9%) 94 (13.4%)

0.13

1425 (18.0%) 62 (20.7%)

1423 (18.0%) 42 (14.0%)

1084 (13.7%) 41 (13.7%)

715 (9.0%) 25 (8.4%)

876 (11.1%) 36 (12.0%)

0.53

1447 (18.1%) 40 (16.9%)

1433 (17.9%) 32 (13.6%)

1097 (13.7%) 28 (11.9%)

712 (8.9%) 28 (11.9%)

881 (11.0%) 31 (13.1%)

0.24

1485 (18.1%) 2 (13.3%)

1464 (17.8%) 1 (6.7%)

1124 (13.7%) 1 (6.7%)

735 (9.0%) 5 (33.3%)

908 (11.1%) 4 (26.7%)

0.006

1459 (18.1%) 28 (18.5%)

1431 (17.7%) 34 (22.5%)

1111 (13.8%) 14 (9.3%)

728 (9.0%) 12 (7.9%)

889 (11.0%) 23 (15.2%)

0.19

65–70

Pulse rate categories at lead-in baseline (−6 weeks) POE N 2634 (30.1%) 1621 (18.5%) Y 294 (28.0%) 186 (17.7%) CVD N 2812 (30.1%) 1724 (18.4%) Y 116 (25.7%) 83 (18.4%) MI N 2821 (29.9%) 1740 (18.4%) Y 107 (29.8%) 67 (18.7%) RCA N 2923 (29.9%) 1806 (18.5%) Y 5 (27.8%) 1 (5.6%) STK N 2862 (29.9%) 1772 (18.5%) Y 66 (29.7%) 35 (15.8%) Pulse rate categories at 12 months POE N 2281(30.3%) Y 212 (30.2%) CVD N 2400 (30.3%) Y 93 (31.1%) MI N 2416 (30.3%) Y 77 (32.6%) RCA N 2491 (30.4%) Y 2 (13.3%) STK N 2453 (30.4%) Y 40 (26.5%)

post randomization

Abbreviations: CVDs, cardiovascular disease; MI, Myocardial infarction; N, no; POE, primary outcome event; RCA, resuscitated cardiac arrest; STK, stroke; Y, yes. a P-value from χ2-test for categorical variables. Event numbers for the component RCA are small 8 thus care is cautioned in interpreting the χ2-test for homogeneity.

(Table 3A) or over the lead-in period or during lead-in baseline to 12 months post randomization (Table 3B). More detailed analyses of the incremental changes showed that in all the subjects without cardiac arrhythmia who had increased their pulse rate by 47.5– ⩽ 10 bpm from lead-in baseline to randomization (that is, when all subjects were receiving sibutramine) the risk of POE was increased (HR: 1.77 CI: 1.07–2.91, © 2015 Macmillan Publishers Limited

P = 0.03). However, this effect was not present when considering pulse rate changes from lead-in baseline to 12 months post randomization (Figure 4). There also appeared to be a trend for those subjects with cardiac arrhythmia who had a decreased pulse of 4 − 2.5 to ⩽ 0 bpm from lead-in baseline to randomization to have an increased risk of POE (HR: 1.86 CI: 0.98–3.51, P = 0.06). For subjects who decreased their pulse from lead-in baseline to International Journal of Obesity (2015) 1 – 9

Sibutramine, weight loss, pulse rate and cardiovascular outcomes RV Seimon et al

6 Table 3B. Primary outcome event components by change categories in pulse rate during lead-in baseline (-6 weeks) to randomization and during lead-in baseline to 12 months o − 10 Change in POE N Y CVDs N Y MI N Y RCA N Y STK N Y

− 10 to − 7.5 − 7.5 to − 5

− 5 to − 2.5

− 2.5 to 0

0–2.5

2.5–5

5–7.5

7.5–10

⩾ 10

P-valuea

pulse rate categories during lead-in baseline to randomization 561 (6.4%) 74 (7.0%)

325 (3.7%) 58 (5.5%)

553 (6.3%) 70 (6.7%)

831 (9.5%) 111 (10.6%)

1411 (16.1%) 1328 (15.2%) 1311 (15.0%) 153 (14.6%) 154 (14.7%) 158 (15.0%)

971 (11.1%) 113 (10.8%)

760 (8.7%) 74 (7.0%)

698 (8.0%) 86 (8.2%)

0.11

605 (6.5%) 30 (6.6%)

351 (3.8%) 32 (7.1%)

595 (6.4%) 28 (6.2%)

899 (9.6%) 43 (9.5%)

1489 (15.9%) 1424 (15.2%) 1397 (14.9%) 1043 (11.2%) 75 (16.6%) 58 (12.8%) 72 (15.9%) 41 (9.1%)

797 (8.5%) 37 (8.2%)

748 (8.0%) 36 (8.0%)

0.07

609 (6.5%) 26 (7.2%)

366 (3.9%) 17 (4.7%)

599 (6.3%) 24 (6.7%)

903 (9.6%) 39 (10.9%)

1514 (16.0%) 1423 (15.1%) 1413 (15.0%) 1045 (11.1%) 50 (13.9%) 59 (16.4%) 56 (15.6%) 39 (10.9%)

811 (8.6%) 23 (6.4%)

758 (8.0%) 26 (7.2%)

0.80

634 (6.5%) 1 (5.6%)

383 (3.9%) 0 (0.0%)

622 (6.4%) 1 (5.6%)

940 (9.6%) 2 (11.1%)

1563 (16.0%) 1479 (15.1%) 1464 (15.0%) 1083 (11.1%) 1 (5.6%) 3 (16.7%) 5 (27.8%) 1 (5.6%)

833 (8.5%) 1 (5.6%)

781 (8.0%) 3 (16.7%)

0.70

618 (6.5%) 17 (7.7%)

374 (3.9%) 9 (4.1%)

606 (6.3%) 17 (7.7%)

915 (9.6%) 27 (12.2%)

1537 (16.0%) 1448 (15.1%) 1444 (15.1%) 1052 (11.0%) 27 (12.2%) 34 (15.3%) 25 (11.3%) 32 (14.4%)

821 (8.6%) 13 (5.9%)

763 (8.0%) 21 (9.5%)

0.24

Change in pulse rate categories from lead-in baseline to POE N 941 (12.5%) 450 (6.0%) 672 (8.9%) Y 92 (13.1%) 54 (7.7%) 68 (9.7%) CVDs N 994 (12.5%) 482 (6.1%) 708 (8.9%) Y 39 (13.0%) 22 (7.4%) 32 (10.7%) MI N 1006 (12.6%) 486 (6.1%) 719 (9.0%) Y 27 (11.4%) 18 (7.6%) 21 (8.9%) RCA N 1032 (12.6%) 503 (6.1%) 738 (9.0%) Y 1 (6.7%) 1 (6.7%) 2 (13.3%) STK N 1008 (12.5%) 491 (6.1%) 727 (9.0%) Y 25 (16.6%) 13 (8.6%) 13 (8.6%)

12 months post randomization 710 (9.4%) 69 (9.8%)

1039 (13.8%) 92 (13.1%)

867 (11.5%) 73 (10.4%)

787 (10.5%) 59 (8.4%)

617 (8.2%) 43 (6.1%)

472 (6.3%) 50 (7.1%)

966 (12.8%) 101 (14.4%)

0.18

748 (9.4%) 31 (10.4%)

1088 (13.7%) 43 (14.4%)

909 (11.5%) 31 (10.4%)

823 (10.4%) 23 (7.7%)

640 (8.1%) 20 (6.7%)

501 (6.3%) 21 (7.0%)

1030 (13.0%) 37 (12.4%)

0.79

754 (9.4%) 25 (10.6%)

1101 (13.8%) 30 (12.7%)

920 (11.5%) 20 (8.5%)

818 (10.2%) 28 (11.9%)

646 (8.1%) 14 (5.9%)

505 (6.3%) 17 (7.2%)

1031 (12.9%) 36 (15.3%)

0.67

778 (9.5%) 1 (6.7%)

1130 (13.8%) 1 (6.7%)

938 (11.4%) 2 (13.3%)

846 (10.3%) 0 (0.0%)

659 (8.0%) 1 (6.7%)

521 (6.3%) 1 (6.7%)

1062 (12.9%) 5 (33.3%)

0.55

767 (9.5%) 12 (7.9%)

1113 (13.8%) 18 (11.9%)

920 (11.4%) 20 (13.2%)

838 (10.4%) 8 (5.3%)

652 (8.1%) 8 (5.3%)

511 (6.3%) 11 (7.3%)

1044 (12.9%) 23 (15.2%)

0.30

Abbreviations: CVDs, cardiovascular disease; MI, Myocardial infarction; N, no; POE, primary outcome event; RCA, resuscitated cardiac arrest; STK, stroke; Y, yes. a P-value from chi square test for categorical variables. Event numbers for the component RCA are small thus care is cautioned in interpreting the χ2-test for homogeneity.

12 months post randomization by 4 − 5 to ⩽ − 2.5 bpm, there was a trend for increased risk (HR: 2.37 CI: 0.97–5.81, P = 0.06); for a pulse decrease of 4 − 10 to ⩽ 7.5 bpm, risk of POE was significantly elevated (HR: 4.08 CI: 1.02–16.32, P = 0.047; Figure 4). It should be noted that investigation into the possible effect modification showed that the effect of sibutramine on POEs was consistent across the pulse rate change category subgroups (no significant interaction effect was found between the pulse rate change and sibutramine). Detailed analysis showing participants in different pulse rate categories at 12 months and changes in pulse rate categories from lead-in baseline to 12 months by various comorbidities (age, β-blockade, baseline cardiac arrhythmias, gender and treatment) are presented in Supplementary Tables S3 and S4. At lead-in baseline, approximately two thirds of the SCOUT population were on β-blocker therapy (N = 6017). When adjusted for treatment, country, age and sex, analyses of β-blocker users in the overall subject population showed a significant influence on POE (HR: 1.17, CI: 1.02–1.33, P = 0.02) and myocardial infarction (HR: 1.36, CI: 1.08–1.72, P = 0.01), when compared with nonusers. Further analyses showed that β-blocker use in subjects without cardiac arrhythmias had no significant influence on POE (HR: 1.07, CI: 0.8–1.45, P = 0.65), whereas in those with cardiac arrhythmias β-blocker use had a trend to increase POE (HR: 1.16, CI: 1.00–1.34, P = 0.06), when compared with nonusers. At lead-in baseline, roughly, 17% of the SCOUT population was ⩾ 70 years of age (N = 1701); those with cardiac arrhythmia, had an increased risk of POE, (HR: 1.35, CI: 1.02–1.79, P = 0.04) as had those without arrhythmia (HR: 1.66, CI: 1.41–1.96, P o 0.0001) when compared with subjects o 70 years of age. International Journal of Obesity (2015) 1 – 9

DISCUSSION We report the surprising observations that in the obese high cardiovascular risk population recruited into the SCOUT trial, intentional weight loss did not influence pulse rate, and that a higher pulse rate during up to 5 years of weight management may be beneficial, which adds new information to the complex issue of heart rate control in cardiovascular subjects without clinical heart failure. Substantial evidence suggests that pulse rate is a key determinant of cardiac ischemia23 and a strong and independent predictor of cardiovascular events in healthy men and women,24 as well as those with established coronary artery disease or previous myocardial infarction.3,25 Furthermore, low pulse rate variability is also a risk for cardiovascular events in populations with26 and without27 known CVDs. The relationship with obesity has not specifically been evaluated, and is likely to be confounded by other obesity-related cardiovascular risk factors28 such as hypertension, heart failure, insulin resistance, diabetes, sleep apnea, inflammation, activated sympathetic nervous system, physical fitness and concomitant drug treatment such as β-blockers. Neither have ‘optimal’ pulse rates for the obese been established. The interaction between obesity, pulse rate and hypertension (a major risk factor for cardiovascular events) is complex. In the Hypertension and Ambulatory Recording Venetia Study of participants who were never treated, both baseline clinic pulse rate (P = 0.007) and 24-h ambulatory pulse rate (P = 0.013), were independent predictors of weight change at study end. In addition, changes in HR during the follow-up either measured in the clinic (P = 0.036) or with 24-h recording (P = 0.009) were © 2015 Macmillan Publishers Limited

Sibutramine, weight loss, pulse rate and cardiovascular outcomes RV Seimon et al

7

Figure 4. Subgroup analysis looking at the effect of treatment on primary outcome events: stratified by changes in pulse rate during lead-in baseline (−6 weeks) to randomization (a and b) and from lead-in baseline to 12 months (c and d), in subjects with (interaction P-value = 0.9) and without (interaction P-value = 0.3) cardiac arrhythmias. Estimates of hazard ratios (95% CI) and log-rank P-values were calculated using the Cox model, adjusted for various factors. Pulse categories: ⩽ − 10, 4 − 10 to ⩽ − 7.5, 4 − 7.5 to ⩽ − 5, 4 − 5.0 to ⩽ − 2.5, 4 − 2.5 to ⩽ 0, 40 to ⩽ 2.5, 42.5 to ⩽ 5, 45.0 to ⩽ 7.5, 47.5 to ⩽ 10 and 410 bpm. CA, cardiac arrhythmia.

independent associates of weight gain.29 Furthermore, in this study, clinic pulse rate and changes in the pulse rate during the first 6 months of follow-up were independent predictors of subsequent systolic blood pressure and diastolic blood pressure regardless of initial blood pressure and other confounders (all P o0.01).30 The impact of pulse rate on cardiovascular outcomes has become of key importance in regulatory assessments of the risks and benefits of pharmacological treatments of obesity. In 2010, the US Food and Drugs Administration,31 and more recently, the European CHMP (Committee on Human Medicinal Products) rejected the approval of a combination of phentermine+topiramate in part because even though there was no overall signal of an increased risk of cardiovascular events in the studies, the consequences of an increased pulse rate in subjects with history of, or with ongoing CVDs, are unknown,32 especially after the recent results of SIGNIFY6 which showed that a lower heart rate is not always better. Glucagon-like peptide 1 (GLP-1) agonists such as exenatide and liraglutide are extensively used for treating type 2 diabetes. A recent systematic review confirms GLP-1 agonists are associated with pulse rate increases of 1.86 bpm (95% CI 0.85–2.87) versus placebo and 1.90 bpm (95% CI 1.30–2.50) versus active diabetes pharmacotherapy control.33 Liraglutide at a higher dose than that used for diabetes treatment (3.0 compared with 0.6–1.8 mg daily) is currently under review by the European CHMP and has also shown increases in pulse rate (mean increase from randomization —after initial weight loss of 3.6+9.4 bpm compared with placebo 2.4+8.6 bpm)34 and in an earlier Phase 2 trial,35 3.5 bpm placebosubtracted. Clinical outcome trials underway, now effectively a © 2015 Macmillan Publishers Limited

prerequisite for regulatory approval, will eventually determine whether these pulse rate changes will offset the welldemonstrated cardioprotective effects of GLP-1 agonists.36–38 Interestingly enough, the correlation between pulse rate and mortality in the Heart Association Study was U-shaped—since excess mortality was found also for pulse rate values o60 bpm.28 This was, in fact, until SIGNIFY, the only study where bradycardia was associated with increased coronary mortality. In the current study, those who had a pre-existing cardiac arrhythmia in whom the pulse rate fell over the course of the study were at an increased risk of POE (HR ranging from 1.86–4.08 with decreasing pulse rate). While in this study those in the sibutramine group had an increase in pulse rate, the increase may in part have been driven by the up-titration of sibutramine dose from 10 to 15 mg that applied to 33% of subjects on sibutramine and 40% in the placebo group during the early post-randomization period as dictated in the protocol in an effort to achieve weight loss. The incidence of POEs was generally lower in subjects who received 15 mg dose than in those who did not across both the treatment groups. Those subjects treated with sibutramine who lost weight showed a graded decrease in pulse rate, compared with those who did not lose weight. In this elderly obese population pulse rate seems to be neither an important modifier nor does it appear to be a clinically useful predictor of the outcomes in an individual subject, at least not compared with the impact of other known cardiovascular markers (for example blood pressure, cholesterol and blood glucose). In this population of obese individuals with CVD there is an unexpected finding that a continuously lower pulse rate had no International Journal of Obesity (2015) 1 – 9

Sibutramine, weight loss, pulse rate and cardiovascular outcomes RV Seimon et al

8 beneficial effect compared with a significantly higher pulse rate, even when the latter was drug induced. Study limitations Assessment of the effects of weight loss and blood pressure management on adverse cardiovascular outcomes was preplanned; however, the detailed choice of categories in the current analyses are post hoc. The SCOUT study population was not typical of a more general population being treated for hypertension; it comprised an older, overweight/obese CVD cohort receiving intensive cardiovascular risk factor monitoring and optimization with multiple adjustable concomitant pharmacotherapies/interventions. We did not adjust for dosage (because of the lack of data) or class of anti-hypertensive agents received or for other potential confounders, especially those which are predictors of poor health, socioeconomic status, job stress or mental health. SCOUT did not assess the presence or severity of possible diabetic-related cardiac autonomic neuropathy, which could modify heart rate responses to sibutramine. The absence of a true placebo group during this initial treatment period (that is, all subjects had an initial 6-week treatment period with sibutramine) prevented the assessment of unequivocal sibutramine-related effects. The changes reported, specifically those relating to pulse rate, may also in part reflect a regression to the mean and this should be taken into account when interpreting the data. CONFLICT OF INTEREST NF: member of SCOUT Executive Steering Committee (ESC) received payments from Abbott, as advisor for Novo Nordisk, Merck, sanofi-aventis, GlaxoSmithKline and Shionogi, consultant for Ajinomoto, and provided expert testimony for sanofi-aventis, Vivus, Arena and received a grant from GlaxoSmithKline. WPTJ: Chair of the SCOUT ESC received payment from Abbott; the International Association for the Study of Obesity when he was the President also received a grant from Novo Nordisk. UFL: was an employee of Abbott with equity interest in the Company. WC: member of the SCOUT ESC received payments from Abbott Laboratories, lecture fees and/or travel reimbursement from Abbott, Ache Laboratorios Farmaceuticos S/A, Roche and Novo Nordisk, as advisor for Abbott, Ache Laboratorios Farmaceuticos S/A, GSK, Novo Nordisk, Takeda and Roche, and provided expert testimony for Abbott. AMS: member of the SCOUT ESC received payment from Abbott Laboratories, and as advisor for Abbott, Merck, Arena, Novo Nordisk, sanofi-aventis, GlaxoSmithKline, Boehringer Ingelheim, and NeuroSearch, as consultant for Vivus and Allegan, provided expert testimony for GlaxoSmithKline, received grants from Abbott and Covidian. LVG: Received a research grant from National Research Funds, Belgium; served on the speaker's bureaus of Sanofi-Aventis and Abbott; served as a consultant to Amylin Pharmaceuticals, Sanofi-Aventis, Eli Lilly, and Abbott; member of SCOUT ESC receiving payment from Abbott. IDC: member of SCOUT ESC received payment from Abbott Laboratories, royalties from Wiley-Blackwell as co-editor of an obesity textbook, and the Boden Institute of Obesity, Nutrition and Exercise received grants from Novo Nordisk, sanofi-aventis, Pfizer (Australia), Weight Watchers, Allergan and the Korean Ministry of Agriculture. The remaining authors declare no conflict of interest.

ACKNOWLEDGEMENTS The original trial was supported and funded by Abbott Laboratories (Abbott Park, IL, USA). The Executive Steering Committee designed the study in cooperation with the sponsor. RVS was supported by a National Health and Medical Research Council of Australia Early Career Research Fellowship (no 1072771). We are also grateful to the Endocrine Society of Australia for a Postdoctoral Award to RVS.

DISCLAIMER The authors have full access to all data, determined the analyses, are solely responsible for its interpretation and this manuscript without reference to the original sponsor and took the final decision to submit the manuscript for publication.

International Journal of Obesity (2015) 1 – 9

REFERENCES 1 Mancia G, Laurent S, Agabiti-Rosei E, Ambrosioni E, Burnier M, Caulfield MJ et al. Reappraisal of European guidelines on hypertension management: a European Society of Hypertension Task Force document. Blood pressure 2009; 18: 308–347. 2 Kannel WB, Sorlie P. Hypertension in Framingham. In: Paul O (ed). Epidemiology and Control of Hypertension. Stratton Intercontinental Medical Book Corp: New York, NY, USA, 1975, pp 553–592. 3 Diaz A, Bourassa MG, Guertin MC, Tardif JC. Long-term prognostic value of resting heart rate in patients with suspected or proven coronary artery disease. Eur Heart J 2005; 26: 967–974. 4 Kolloch R, Legler UF, Champion A, Cooper-Dehoff RM, Handberg E, Zhou Q et al. Impact of resting heart rate on outcomes in hypertensive patients with coronary artery disease: findings from the INternational VErapamil-SR/trandolapril STudy (INVEST). Eur Heart J 2008; 29: 1327–1334. 5 Fox K, Ford I, Steg PG, Tendera M, Robertson M, Ferrari R. Heart rate as a prognostic risk factor in patients with coronary artery disease and left-ventricular systolic dysfunction (BEAUTIFUL): a subgroup analysis of a randomised controlled trial. Lancet 2008; 372: 817–821. 6 Fox K, Ford I, Steg PG, Tardif JC, Tendera M, Ferrari R. Ivabradine in stable coronary artery disease without clinical heart failure. N Engl J Med 2014; 371: 1091–1099. 7 Swedberg K, Komajda M, Bohm M, Borer JS, Ford I, Dubost-Brama A et al. Ivabradine and outcomes in chronic heart failure (SHIFT): a randomised placebocontrolled study. Lancet 2010; 376: 875–885. 8 Fox K, Komajda M, Ford I, Robertson M, Bohm M, Borer JS et al. Effect of ivabradine in patients with left-ventricular systolic dysfunction: a pooled analysis of individual patient data from the BEAUTIFUL and SHIFT trials. Eur Heart J 2013; 34: 2263–2270. 9 Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo JL Jr et al. The seventh report of the Joint National Committee on prevention, detection, evaluation, and treatment of high blood pressure: the JNC 7 report. JAMA 2003; 289: 2560–2572. 10 Graham I, Atar D, Borch-Johnsen K, Boysen G, Burell G, Cifkova R et al. European guidelines on cardiovascular disease prevention in clinical practice: executive summary: Fourth Joint Task Force of the European Society of Cardiology and Other Societies on Cardiovascular Disease Prevention in Clinical Practice (Constituted by representatives of nine societies and by invited experts). Eur Heart J 2007; 28: 2375–2414. 11 Snow V, Barry P, Fitterman N, Qaseem A, Weiss K. Pharmacologic and surgical management of obesity in primary care: a clinical practice guideline from the American College of Physicians. Ann Intern Med 2005; 142: 525–531. 12 Lean M, Finer N. ABC of obesity management: part II--drugs. BMJ 2006; 333: 794–797. 13 Halford JC, Boyland EJ, Cooper SJ, Dovey TM, Huda MS, Dourish CT et al. The effects of sibutramine on the microstructure of eating behaviour and energy expenditure in obese women. J Psychopharmacol 2010; 24: 99–109. 14 Heusser K, Tank J, Diedrich A, Engeli S, Klaua S, Kruger N et al. Influence of sibutramine treatment on sympathetic vasomotor tone in obese subjects. Clin Pharmacol Ther 2006; 79: 500–508. 15 Lechin F, van der Dijs B, Hernandez G, Orozco B, Rodriguez S, Baez S. Neurochemical neuroautonomic and neuropharmacological acute effects of sibutramine in healthy subjects. Neurotoxicology 2006; 27: 184–191. 16 James WP, Caterson ID, Coutinho W, Finer N, Van Gaal LF, Maggioni AP et al. Effect of sibutramine on cardiovascular outcomes in overweight and obese subjects. N Engl J Med 2010; 363: 905–917. 17 Seimon RV, Espinoza D, Ivers L, Gebski V, Finer N, Legler UF et al. Changes in body weight and blood pressure: paradoxical outcome events in overweight and obese subjects with cardiovascular disease. Int J Obes (Lond) 2014; 38: 1165–1171. 18 James WPT. The SCOUT study: risk-benefit profile of sibutramine in overweight high-risk cardiovascular patients. Eur Heart J Suppl 2005; 7: L44–L48. 19 Rosendorff C, Black HR, Cannon CP, Gersh BJ, Gore J, Izzo JL Jr et al. Treatment of hypertension in the prevention and management of ischemic heart disease: a scientific statement from the American Heart Association Council for High Blood Pressure Research and the Councils on Clinical Cardiology and Epidemiology and Prevention. Circulation 2007; 115: 2761–2788. 20 Kumanyika SK, Obarzanek E, Stettler N, Bell R, Field AE, Fortmann SP et al. Population-based prevention of obesity: the need for comprehensive promotion of healthful eating, physical activity, and energy balance: a scientific statement from American Heart Association Council on Epidemiology and Prevention, Interdisciplinary Committee for Prevention (formerly the expert panel on population and prevention science). Circulation 2008; 118: 428–464. 21 Furie KL, Kasner SE, Adams RJ, Albers GW, Bush RL, Fagan SC et al. Guidelines for the prevention of stroke in patients with stroke or transient ischemic attack: a guideline for healthcare professionals from the american heart association/ american stroke association. Stroke 2011; 42: 227–276.

© 2015 Macmillan Publishers Limited

Sibutramine, weight loss, pulse rate and cardiovascular outcomes RV Seimon et al

9 22 Ryden L, Standl E, Bartnik M, Van den Berghe G, Betteridge J, de Boer MJ et al. Guidelines on diabetes, pre-diabetes, and cardiovascular diseases: executive summary. The Task Force on Diabetes and Cardiovascular Diseases of the European Society of Cardiology (ESC) and of the European Association for the Study of Diabetes (EASD). Eur Heart J 2007; 28: 88–136. 23 Andrews TC, Fenton T, Toyosaki N, Glasser SP, Young PM, MacCallum G et al. Subsets of ambulatory myocardial ischemia based on heart rate activity. Circadian distribution and response to anti-ischemic medication. Circulation 1993; 88: 92–100. 24 Cooney MT, Vartiainen E, Laatikainen T, Juolevi A, Dudina A, Graham IM. Elevated resting heart rate is an independent risk factor for cardiovascular disease in healthy men and women. Am Heart J 2010; 159: 612–9 e3. 25 Fox K, Borer JS, Camm AJ, Danchin N, Ferrari R, Lopez Sendon JL et al. Resting heart rate in cardiovascular disease. J Am Coll Cardiol 2007; 50: 823–830. 26 Liao D, Carnethon M, Evans GW, Cascio WE, Heiss G. Lower heart rate variability is associated with the development of coronary heart disease in individuals with diabetes: the atherosclerosis risk in communities (ARIC) study. Diabetes 2002; 51: 3524–3531. 27 Hillebrand S, Gast KB, de Mutsert R, Swenne CA, Jukema JW, Middeldorp S et al. Heart rate variability and first cardiovascular event in populations without known cardiovascular disease: meta-analysis and dose-response meta-regression. Europace 2013; 15: 742–749. 28 Dyer AR, Persky V, Stamler J, Paul O, Shekelle RB, Berkson DM et al. Heart rate as a prognostic factor for coronary heart disease and mortality: findings in three Chicago epidemiologic studies. Am J Epidemiol 1980; 112: 736–749. 29 Palatini P, Mos L, Santonastaso M, Zanatta N, Mormino P, Saladini F et al. Resting heart rate as a predictor of body weight gain in the early stage of hypertension. Obesity (Silver Spring) 2011; 19: 618–623.

30 Palatini P, Dorigatti F, Zaetta V, Mormino P, Mazzer A, Bortolazzi A et al. Heart rate as a predictor of development of sustained hypertension in subjects screened for stage 1 hypertension: the HARVEST Study. J Hypertens 2006; 24: 1873–1880. 31 US Food and Drug Administration. Proceedings of the Endocrinologic and Metabolic Drugs Advisory Committee Meeting. 1995: Silver Spring, MD: US Food and Drug Administration; 1995. 32 European Medicines Agency. Scientific conclusions and grounds for refusal. Available online http://wwwemaeuropaeu/docs/en_GB/document_library/Other/ human/002350/WC500147577pdf. 33 Robinson LE, Holt TA, Rees K, Randeva HS, O'Hare JP. Effects of exenatide and liraglutide on heart rate, blood pressure and body weight: systematic review and meta-analysis. BMJ open 2013; 3: e001986. 34 Wadden TA, Hollander P, Klein S, Niswender K, Woo V, Hale PM et al. Weight maintenance and additional weight loss with liraglutide after low-caloriediet-induced weight loss: the SCALE maintenance randomized study. Int J Obes (Lond) 2013; 37: 1443–1451. 35 Astrup A, Carraro R, Finer N, Harper A, Kunesova M, Lean ME et al. Safety, tolerability and sustained weight loss over 2 years with the once-daily human GLP-1 analog, liraglutide. Int J Obes (Lond) 2012; 36: 843–854. 36 Myat A, Redwood SR, Gersh BJ, Yellon DM, Marber MS. Diabetes, incretin hormones and cardioprotection. Heart 2014; 100: 1550–1561. 37 Picatoste B, Ramirez E, Caro-Vadillo A, Iborra C, Ares-Carrasco S, Egido J et al. Sitagliptin reduces cardiac apoptosis, hypertrophy and fibrosis primarily by insulin-dependent mechanisms in experimental type-II diabetes. Potential roles of GLP-1 isoforms. PloS One 2013; 8: e78330. 38 Wang D, Luo P, Wang Y, Li W, Wang C, Sun D et al. Glucagon-like peptide-1 protects against cardiac microvascular injury in diabetes via a cAMP/PKA/ Rho-dependent mechanism. Diabetes 2013; 62: 1697–1708.

Supplementary Information accompanies this paper on International Journal of Obesity website (http://www.nature.com/ijo)

© 2015 Macmillan Publishers Limited

International Journal of Obesity (2015) 1 – 9