European Journal of Preventive Cardiology

Piepoli, Massimo F

doi: 10.1177/2047487320936413pmid: 32583735

The heart in the time of COVID-19 Coronavirus disease 2019 (COVID-19) has been completely upsetting our society so that every kind of daily behaviour had to be adapted and modified; this is evident for scientific and medical communication too. Here, authors from Italy and Spain, the two countries in Europe that have been most hit by COVID-19, have promoted some reflections for every physician professionally involved both in the war against COVID-19 and in treating ‘traditional’ patients without this new disease. We should rethink our clinical management and routine controls of different cardiovascular diseases (e.g. heart failure, chronic coronary syndromes) in favour of out-of-hospital diffused and programmed assistance (including e-health)1 to reduce the risk of infection and guarantee the same level of care to all our patients. Efficacy of extended, comprehensive outpatient cardiac rehabilitation The importance and the challenge of long-term maintenance of risk factor control after acute events are well evident.2 In this regard, the benefit of an intensive long-term (12 months) prevention programme after acute myocardial infarction using personal teachings and telemetric strategies has recently been published.3 Here a new Austrian model of long-term, phase III secondary prevention lasting 6–12 months is presented, in which endurance and strength training sessions were carried out regularly and lasted up to 50 minutes each. A large cohort of consecutive patients from a national registry significantly improved their metabolic risk factor profile and increased exercise capacity. Adequate and accredited programmes need to be established nationwide to help patients comply with medical recommendations of lifelong lifestyle changes.4 Walking pace improves the prediction of mortality Current established risk scores for cardiovascular or all-cause mortality largely rely on non-modifiable (age, sex) or biological (e.g. blood pressure, cholesterol) risk factors. The discordance between health promotion campaigns, which are mainly based on modifiable lifestyle behaviours (i.e. encouragement of physical activity) and risk prediction is therefore evident. Consequently, behaviour change does not necessarily affect current risk prediction, while risk prediction does not reinforce the importance of healthy lifestyle behaviours. A UK Biobank prognostic study aimed to quantify and rank systematically the potential usefulness of simple, easily collected dietary, physical activity and physical function variables as prognostic markers for mortality in comparison with, and when added to, the SCORE risk factors. Importantly, walking pace was found to improve risk prediction but no other indicators. This is in contrast with other studies demonstrating the benefit of healthy diet on lowering cardiovascular risk.5 A study limitation was that risk factor control was assessed by self-reported questionnaires. Cost-effectiveness of exercise therapy There is strong evidence for the effectiveness of prescribed supervised exercise therapy, exercise training and exercise-based cardiac rehabilitation in patients with cardiovascular diseases,6,7 but economic evaluations are limited. A systematic search of seven electronic databases identified 15 economic evaluations conducted alongside prescribed supervised exercise therapy, showing that intervention was highly cost-effective in coronary heart disease, chronic heart failure, intermittent claudication, body mass index greater than 25 kg/m2, but with less evidence in hypertension or type 2 diabetes mellitus. Non-alcoholic fatty liver disease and cardiovascular disease Non-alcoholic fatty liver disease (NAFLD) is the leading cause of chronic liver disease worldwide, and it encompasses a wide spectrum of conditions, ranging from simple steatosis to non-alcoholic steatohepatitis, fibrosis and hepatocellular carcinoma. Although NAFLD is associated with an increased risk of liver-related morbidity or mortality, it is now considered a multisystem disorder, which affects a variety of extra-hepatic organs, including the cardiovascular system. In this context, here an ‘In the news.’ article claimed that NAFLD patients should be screened for cardiovascular diseases on a regular basis. It is not surprising that NAFLD associates with an increased risk, because they share common risk factors such as abdominal obesity, hypertension, atherogenic dyslipidemia and insulin resistance/dysglycemia. In addition, multiple common pathophysiological mechanisms may play a role, such as systemic inflammation, endothelial dysfunction, oxidative stress and the role of pro-protein convertase subtilisin/kexin type 9 (PCSK9).8 Septal myectomy in hypertrophic cardiomyopathy: predictors of improved exercise capacity Hypertrophic cardiomyopathy (HCM) is a prevalent inherited heart disease that affects one in 500 individuals and, importantly, 70% of these patients exhibit the hypertrophic obstructive phenotype, with greater risk of mortality, and more progressive heart failure severity.9 Septal myectomy surgery is the gold standard treatment for patients who do not respond to medical therapy, intervention that leads to increased long-term survival, and exercise capacity (VO2peak). Here a large cohort of patients with HCM demonstrated that demographic (i.e. female sex, age), lack of cardiac rehabilitation enrolment and cardiovascular risk factors (i.e. history of dyslipidemia) were predictive of those HCM patients who did not exhibit improvements in exercise capacity following septal myectomy surgery. These findings demonstrate the importance of individual clinical characteristics influencing peak exercise capacity following septal myectomy surgery in patients with HCM. Suboptimal health behaviours in grown-up congenital heart disease Optimal health behaviours are essential in the maintenance of good health and reduction of risk for cardiovascular complications, particularly in adults with grown-up congenital heart disease, because they are at higher risk than the general population for cardiovascular events. A large international registry has shown that a substantial percentage did not follow optimal health behaviours (i.e. 10% binge drinking, 12% cigarette smoking and 6% at least monthly use of recreational drugs). Moreover, despite the emphasis placed during clinic appointments on dental hygiene, aimed at minimising the risk of infective endocarditis, over a quarter of patients fail to see their dentist annually and almost one-third brush their teeth less than twice daily. Finally, despite moving away from exercise restriction for most patients, fewer than half participate in sports. The findings are somewhat disconcerting considering that these are mostly patients under regular follow-up in specialised centres, who meet health professionals regularly throughout their lives. The results of this study can be interpreted as our failure as healthcare providers to educate patients and encourage behaviour associated with a healthy lifestyle.10 Moreover, a system is required that targets patients who are more likely to exhibit at-risk behaviours, providing them with sufficient information and stimuli to maximise engagement.1,11 Remnant cholesterol and residual cardiovascular risk Despite achieving optimal low-density lipoprotein cholesterol levels, the residual cardiovascular risk has been attributed to triglyceride-rich lipoproteins and their cholesterol content, known as remnant cholesterol (RC). RC has gained increasing recognition as a biomarker driving residual risk in this contemporary era of greater obesity, diabetes and metabolic syndrome rates. Here in a large database involving 5754 patients, RC was significantly associated with coronary atheroma progression, regardless of biochemical and clinical risk factors. This suggests that accelerated progression of atherosclerosis is an important factor underlying the observation of a greater incidence of clinical cardiovascular events. Measuring RC is likely to play an important role in identifying patients requiring more intense or personalised medical therapy for secondary prevention. These data also highlight RC and triglyceride targeted therapies as areas of interest for the clinical development of novel anti-atherosclerotic agents. 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 Frederix I , Caiani EG, Dendale P, et al. . ESC e-Cardiology Working Group Position Paper: Overcoming challenges in digital health implementation in cardiovascular medicine . Eur J Prev Cardiol 2019 ; 26 : 1166 – 1177 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Kotseva K , De Backer G, De Bacquer D, et al. . Lifestyle and impact on cardiovascular risk factor control in coronary patients across 27 countries: results from the European Society of Cardiology ESC-EORP EUROASPIRE V registry . Eur J Prev Cardiol 2019 ; 26 : 824 – 835 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Wienbergen H , Fach A, Meyer S, et al. . Effects of an intensive long-term prevention programme after myocardial infarction – a randomized trial . Eur J Prev Cardiol 2019 ; 26 : 522 – 530 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Janssen A , Wagenaar KP, Dendale P, et al. . Accreditation of clinical centres providing primary prevention, secondary prevention and rehabilitation, and sports cardiology: a step towards optimizing quality . Eur J Prev Cardiol 2019 ; 26 : 1775 – 1777 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Kwok CS , Gulati M, Michos ED, et al. . Dietary components and risk of cardiovascular disease and all-cause mortality: a review of evidence from meta-analyses . Eur J Prev Cardiol 2019 ; 26 : 1415 – 1429 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Doyle MP , Indraratna P, Tardo DT, et al. . Safety and efficacy of aerobic exercise commenced early after cardiac surgery: a systematic review and meta-analysis . Eur J Prev Cardiol 2019 ; 26 : 36 – 45 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Saeidifard F , Medina-Inojosa JR, West CP, et al. . The association of resistance training with mortality: a systematic review and meta-analysis . Eur J Prev Cardiol 2019 ; 26 : 1647 – 1665 Google Scholar Crossref Search ADS PubMed WorldCat 8 Macchi C , Banach M, Corsini A, et al. . Changes in circulating pro-protein convertase subtilisin/kexin type 9 levels – experimental and clinical approaches with lipid-lowering agents. Eur J Prev Cardiol 2019 ; 26 : 930 – 949 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Finocchiaro G , Papadakis M, Dhutia H, et al. . Obesity and sudden cardiac death in the young: clinical and pathological insights from a large national registry . Eur J Prev Cardiol 2018 ; 25 : 395 – 401 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Villani GQ , Villani A, Zanni A, et al. . Mobile health and implantable cardiac devices: patients’ expectations. Eur J Prev Cardiol 2019 ; 26 : 920 – 927 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Ariansen I , Strand BH, Kjøllesdal MKR, et al. . The educational gradient in premature cardiovascular mortality: examining mediation by risk factors in cohorts born in the 1930s, 1940s and 1950s . Eur J Prev Cardiol 2019 ; 26 : 1096 – 1103 . Google Scholar Crossref Search ADS PubMed WorldCat © The European Society of Cardiology 2020 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 2020

Sciatti, Edoardo; Ceconi, Claudio

doi: 10.1177/2047487320925622pmid: 32397803

The infection of the novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), namely coronavirus disease 2019 (COVID-19), has recently become a pandemic, spreading worldwide as rapidly as could not have been thought possible before.1 The disease has been completely upsetting our society so that every kind of daily behaviour has been adapted and modified; this is evident for scientific and medical communication too. The most important scientific media are answering the need for rapid circulation of information throughout the world, while the external validation of the details is delegated to subsequent analysis and checking. Experimental verification has become empirical in a hypothetical laboratory as big as the entire world. Moving in this direction, Barison et al. filtered the information that Chinese colleagues had provided through their daily experience and wrote an interesting and pragmatic review.2 The authors come from Italy and Spain, the two countries in Europe that have been most hit by COVID-19. They have the merit of promoting some reflections for every physician who is professionally involved both in the war against COVID-19 and in treating ‘traditional’ patients without this new disease. The context of the review is very complex and articulated, regarding epidemiology, pathophysiology and therapy of COVID-19. Although the review is very complete, in our opinion some issues are worth further comment. First, an obvious difference exists between the prevalence of an associated condition in the affected population and its negative prognostic role. In general, we suggest caution in extrapolating the real role of a prognosticator from meta-analyses. They are subjected to adjustments and corrections which are approximate at most, so that estimating the true role of a factor is arduous. As a provoking example, hypertension is prevalent among COVID-19 patients with unfavourable prognosis,3 but numerically the prevalence is similar to that expected for every age strata in the general population: hypertension is as age-dependent as it is found in COVID-19. Is hypertension really a prognosticator or only an innocent spectator? There is a compelling need for adjusted analyses, which are solely reliable for guiding preventive and therapeutic strategies. The same can be said for the role of a smoking habit. Asians are less prone to smoke and this could have influenced the statistical models, impairing their validity for Western world inhabitants. Second, the authors considered, in a comprehensive way, the most frequent drugs administered to patients affected by cardiovascular (CV) disease in the COVID-19 era: this is very timely, appropriate and clinically useful.2 Regarding renin-angiotensin-aldosterone system (RAAS) inhibitors, the position statement of the European Society of Cardiology4 is clear and mandatory, but the decision to maintain or to withdraw RAAS inhibitors in critical ill patients is left to an individual physician, with consideration of a number of aspects (e.g. blood pressure, kidney injury, dehydration, heart involvement). Third, besides the side effects of the anti-COVID-19 therapies (e.g. QT interval prolongation for chloroquine, hydroxychloroquine and azithromycin), Barison et al. highlight the most important aspect not to forget: there are considerable drug–drug interactions between CV therapies and molecules used in caring for COVID-19 patients.2,5 This is a very relevant issue in everyday clinical practice. Finally, after two months of difficult cohabitation with SARS-CoV-2 in the Western world, we think that the conclusion of the authors is extremely worthy of sharing: Patient management should be redefined based on the risk of viral transmission to the medical staff, the overlap clinical presentation between COVID-19 and more common acute cardiac diseases, as well as the pharmacological interactions between antiviral, anti-inflammatory and CV drugs.2 We consider that cardiologists worldwide should rethink in-hospital management and routine controls of different CV diseases (e.g. heart failure, chronic coronary syndromes) in favour of out-of-hospital diffused and programmed assistance to reduce the risk of infection and guarantee the same level of care to these high-risk patients. Declaration of conflicting interests The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article. Funding The author(s) received no financial support for the research, authorship and/or publication of this article. References 1 Li R , Pei S, Chen B, et al. . Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV2). Science. Epub ahead of print 16 March 2020 . DOI: 10.1126/science.abb3221. 2 Barison A , Aimo A, Castiglione V, et al. . Cardiovascular disease and COVID-19: Les liaisons dangereuses. Eur J Prev Cardiol . 2020 ; 27 : 1017 – 1025 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Yang J , Zheng Y, Gou X, et al. . Prevalence of comorbidities in the novel Wuhan coronavirus (COVID-19) infection: A systematic review and meta-analysis. Int J Infect Dis 2020; 94: 91–95. 4 De Simone G. Position Statement of the ESC Council on Hypertension on ACE-Inhibitors and Angiotensin Receptor Blockers. https://www.escardio.org/Councils/Council-on-Hypertension-(CHT)/News/position-statement-of-the-esc-council-on-hypertension-on-ace-inhibitors-and-ang (2020, accessed 18 April 2020). 5 The Liverpool Drug Interaction Group (based at the University of Liverpool, UK), in collaboration with the University Hospital of Basel (Switzerland) and Radboud UMC (Netherlands). COVID-19 Drug Interactions. https://www.covid19-druginteractions.org/ (2020, accessed 18 April 2020).. © The European Society of Cardiology 2020 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 2020

Barison, Andrea; Aimo, Alberto; Castiglione, Vincenzo; Arzilli, Chiara; Lupón, Josep; Codina, Pau; Santiago-Vacas, Evelyn; Cediel, Germán; Emdin, Michele; Bayes-Genis, Antoni

doi: 10.1177/2047487320924501pmid: 32391719

Reich, Bernhard; Benzer, Werner; Harpf, Hanns; Hofmann, Peter; Mayr, Karl; Ocenasek, Helmuth; Podolsky, Andrea; Pokan, Rochus; Porodko, Michael; Puelacher, Christoph; Sareban, Mahdi; Traninger, Heimo; Ziegelmeyer, Wolfgang; Niebauer, Josef

Jelinek, Michael; Vale, Margarite

doi: 10.1177/2047487319891518pmid: 31795760

In 2013, Hemingway et al. introduced the concept of prognostic research as opposed to aetiological research. The PROGRESS group regard prognostic research as fundamental to clinical and public health research.1–4 Argyridou et al. have applied these concepts in 298,829 subjects from the UK Biobank who were free of cancer and cardiovascular disease at baselines.5 These subjects completed nutritional and physical activity questionnaires at baseline and were followed up for a median of 6.9 years. The investigators evaluated whether or not dietary and physical activity measures at baseline added to the prediction of all-cause and cardiovascular risk calculated from the European Society of Cardiology risk prediction models of prognosis. Their analysis demonstrated that only fast walking speed improved the concordance statistic (C-statistic) for the prediction of all-cause and cardiovascular mortality in men and women when added to the SCORE measure of cardiovascular risk. Dietary cereal intake improved the C-score in men but not in women. Fast walking has been shown to correlate with direct measurement of physical fitness.5 Coronary risk scores used widely throughout the world are aetiological cardiovascular risk scores. All of the risk factor scores use age, sex, cigarette smoking, high blood pressure, dyslipidaemia, and diabetes mellitus in their risk prediction scores. Of these, only cigarette smoking is a measure of lifestyle. The rest are biomedical risk factors. It was many years after these factors were derived from epidemiological studies that their predictive power was confirmed. Indeed, the major predictive factor in these equations using the C-statistic is the chronological age of the subject.6 Dietary and physical activity measures have been associated with the development of cardiovascular disease, cancer, and all-cause mortality.7,8 It may well be that diet and physical activity are primordial risk factors, which generate their prognostic effect chiefly through the conventional cardiovascular risk factors. This has implications for the secondary prevention of cardiovascular disease. The COACH Program – a standardised, structured coaching program delivered by highly trained health professionals to people already diagnosed with a chronic disease or who are at high risk of developing chronic disease – has recently been named the most evidence-based cardiovascular disease prevention program in the world on clinical and cost-effectiveness by the British Heart Foundation and Public Health England’s International Cardiovascular Disease Prevention case studies report, October 2018.9 Delivered by telephone and mail-outs over a period of six months, the COACH Program is focused on closing the “evidence–practice gap” – the gap between guideline-recommended care and the care patients actually receive. The coaches identify the treatment gaps for each individual patient, when medication has not been optimised and risk factor targets for their modifiable risk factors (e.g. lipids, blood pressure, blood glucose and glycated haemoglobin levels, alcohol intake, smoking, physical activity, waist measurement) have not been achieved. The COACH Program informs patients of their specific gaps in treatment and then provides explicit advice on how to close the gaps and achieve national guideline-recommended target levels for their modifiable risk factors while the patients work with their usual doctors. Each verbal coaching session is followed by a structured written report that summarises the session. The COACH Program has been shown to be superior to usual medical care in reducing risk factors in patients with cardiovascular disease in two randomised controlled trials.10 Clinical audits show that the program maintains the improvements long term, achieves greater benefit for socioeconomically disadvantaged people than the more affluent, reaches people in remote locations where face-to-face programs are not feasible, and is as effective in indigenous people as non-indigenous people. The COACH Program was found to significantly lower mortality and reduce payer costs of a private health insurer in 512 patients who were propensity matched with 512 patients eligible but not receiving the COACH Program followed for a mean of 6.35 years.11 This UK Biobank paper supports the proposition that the best lifestyle can reduce the incidence of aetiological risk factors for cardiovascular disease and premature mortality. It also supports our proposition that once cardiovascular disease has occurred, the best secondary prevention approach is to reduce the evidence–practice gap, particularly in the area of biomedical risk factors. It takes many years to develop prognostic risk factors and validated therapies. Surrogate measures that correlate with the prognostic outcomes may be useful in shortening this process. We suggest that the differences between health promotion and the secondary prevention of cardiovascular disease could result in two different measures of surrogate risk. Health promotion programs could be compared using biomedical risk factors as outcomes of their efforts into diet, exercise and fitness, cigarette consumption, and alcohol intake. Conversely, secondary prevention programs could be compared by their impacts on the reduction of the evidence–practice gaps in biomedical risk factors. 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 Hemingway H , Croft P, Perel P, et al. Prognosis research strategy (PROGRESS) 1: a framework for researching clinical outcomes . BMJ 2013 ; 346 : e5595 – e5595 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Riley RD , Hayden JA, Steyerberg EW, et al. Prognosis research strategy (PROGRESS) 2: prognostic factor research . PLoS Med 2013 ; 10(2) : e1001380 – e1001380 . Google Scholar Crossref Search ADS WorldCat 3 Steyerberg EW , Moons KJM, van der Windt DA, et al. Prognosis research strategy (PROGRESS) 3: prognostic model research . PLoS Med 2013 ; 10(2) : e1001381 – e1001381 . Google Scholar Crossref Search ADS WorldCat 4 Hingorani AD , van der Windt DA, Riley RD, et al. Prognosis research strategy (PROGRESS) 4: stratified medicine research . BMJ 2013 ; 346 : e5793 – e5793 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Argyridou S, Zaccardi F, Davies MJ, et al. Walking pace improves all-cause and cardiovascular mortality risk prediction: a UK Biobank prognostic study. Eur J Prev Cardiol. 2020; 27: 1036–1044 . 6 Cook NR . Use and misuse of the receiver operating characteristic curve in risk prediction . Circulation 2007 ; 115 : 928 – 935 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Aune D , Giovannucci E, Boffetta P, et al. Fruit and vegetable intake and the risk of cardiovascular disease, total cancer and all-cause mortality – a systematic review and dose-response meta-analysis of prospective studies . Int J Epidemiol 2017 ; 46 : 1029 – 1046 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Harber MP , Kaminsky LA, Arena R, et al. Impact of cardio-respiratory fitness on all-cause and disease-specific mortality. Advances since 2009 . Prog Cardiovasc Dis 2017 ; 60 ( 1 ): 11 – 20 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Solutions for Public Health, British Heart Foundation, Public Health England. International Cardiovascular Disease Prevention Case studies, October 2018. Available at: https://www.bhf.org.uk/for-professionals/healthcare-professionals/data-and-statistics/international-cardiovascular-disease-case-studies (accessed 5 November 2019) . 10 Vale MJ , Jelinek MV, Best JD, et al. Coaching Patients on achieving cardiovascular health (COACH). A multicentre randomized controlled trial in patients with coronary heart disease . Arch Intern Med 2003 ; 163 : 2775 – 2783 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Byrnes J , Elliott T, Vale MJ, et al. Coaching patients saves life and money . Am J Med 2018 ; 131 : 415–121. Google Scholar OpenURL Placeholder Text WorldCat © The European Society of Cardiology 2020 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 2020

Argyridou, Stavroula; Zaccardi, Francesco; Davies, Melanie J; Khunti, Kamlesh; Yates, Thomas

doi: 10.1177/2047487319887281pmid: 31698963

Aims The purpose of this study was to quantify and rank the prognostic relevance of dietary, physical activity and physical function factors in predicting all-cause and cardiovascular mortality in comparison with the established risk factors included in the European Society of Cardiology Systematic COronary Risk Evaluation (SCORE). Methods We examined the predictive discrimination of lifestyle measures using C-index and R2 in sex-stratified analyses adjusted for: model 1, age; model 2, SCORE variables (age, smoking status, systolic blood pressure, total and high-density lipoprotein cholesterol). Results The sample comprised 298,829 adults (median age, 57 years; 53.5% women) from the UK Biobank free from cancer and cardiovascular disease at baseline. Over a median follow-up of 6.9 years, there were 2174 and 3522 all–cause and 286 and 796 cardiovascular deaths in women and men, respectively. When added to model 1, self-reported walking pace improved C-index in women and men by 0.013 (99% CI: 0.007–0.020) and 0.022 (0.017–0.028) respectively for all-cause mortality; and by 0.023 (0.005–0.042) and 0.034 (0.020–0.048) respectively for cardiovascular mortality. When added to model 2, corresponding values for women and men were: 0.008 (0.003–0.012) and 0.013 (0.009–0.017) for all-cause mortality; and 0.012 (–0.001–0.025) and 0.024 (0.013–0.035) for cardiovascular mortality. Other lifestyle factors did not consistently improve discrimination across models and outcomes. The pattern of results for R2 mirrored those for C-index. Conclusion A simple self-reported measure of walking pace was the only lifestyle variable found to improve risk prediction for all-cause and cardiovascular mortality when added to established risk factors. Walking pace, mortality, cardiovascular risk, prognosis Introduction Cardiovascular disease (CVD) is a global epidemic representing a major public health issue.1 Modifiable lifestyle behaviours, such as smoking, physical inactivity and unhealthy diets are established risk factors for cardiovascular mortality.2–5 Epidemiological evidence further suggests that self-reported or easily collected measures of physical function, such as walking pace and handgrip strength, are also strongly associated with mortality and CVD.6–8 However, while the literature reporting associations among diet, physical activity and physical function with health outcomes is extensive, there has been less attempt to quantify whether these factors improve the prediction of mortality outcomes in a large contemporary cohort, with analysis largely restricted to single lifestyle factors and all-cause mortality.9,10 In recent years, a focused approach on prognostic research has been suggested whereby the identification of an association between risk factors and outcomes represents only the initial research step (i.e. aetiological association).11 Proof of association does not necessarily translate into a better prediction ability, even in the case of very strong, independent and causal associations. Such a relevant distinction relies on the substantial difference between aetiological and prognostic research as the prediction of an outcome is not equivalent to explaining its cause.11 Investigating the prognostic importance of dietary and lifestyle factors also has potential for public health relevance. Current established risk scores for cardiovascular or all-cause mortality largely relay on non-modifiable (age, sex) or biological (e.g. blood pressure, cholesterol) risk factors.12–15 For example, the European Society of Cardiology Systematic Coronary Risk Evaluation (SCORE) risk score, which was developed to estimate the probability of cardiovascular mortality, is based on age, smoking status, systolic blood pressure, total and high-density lipoprotein (HDL) cholesterol.15 SCORE is one of the most widely investigated and validated risk models developed on a European population.16–19 The discordance between health promotion campaigns, which are based on modifiable lifestyle behaviours (i.e. encouragement of physical activity), and risk prediction is therefore evident. Consequently, behaviour change does not necessarily affect current risk prediction, whilst risk prediction does not reinforce the importance of healthy lifestyle behaviours. This limitation has been recognised in the development of prediction tools for some chronic diseases: for example, the established Finnish Diabetes Risk Score included questions related to diet and physical activity for educational purposes and not because those particular questions were found to be prognostically relevant.20 Research is therefore needed across a broad range of lifestyle factors to identify those with the greatest potential for improving risk prediction for disease outcomes, including cardiovascular disease. In view of these limitations and of current standards for defining clinical usefulness of risk factors in risk prediction,21 we aimed to systematically quantify and rank the potential usefulness of simple, easily collected dietary, physical activity and physical function variables as prognostic markers for all-cause and cardiovascular mortality in comparison with, and when added to, the SCORE risk factors in the contemporary UK Biobank population. This will provide novel evidence as to whether lifestyle variables that could easily be collected within routine care settings add to the prognostic information that is already contained within established risk factors. Methods UK Biobank This analysis used data from 502,621 volunteers within UK Biobank, aged between 38–73 years at recruitment (UK Biobank Application Number 3140). UK Biobank is an ongoing prospective cohort study with data collected in 22 centres throughout England, Wales and Scotland between March 2006–July 2010. Data field (DF) identification numbers are reported below for each variable included in this analysis and can be used to search for detailed information about measurement procedures within the UK Biobank data showcase.22 Ethical approval for the UK Biobank study was obtained from the North West Centre for Research Ethics Committee (MREC, 11/NW/0382). In Scotland, UK Biobank has approval from the Community Health Index Advisory Group (CHIAG). Diet variables A touchscreen food frequency questionnaire captured information on diet. We used several diet variables: oily fish (DF 1329), non-oily fish (1339), poultry (1359), cheese (1408), beef (1369), lamb/mutton (1379) and pork (1389) intake, fresh fruit (1309), dried fruit (1319), bread (1438) and cereal (1458). Processed meat (1349) was defined through a single variable assessing any intake of bacon, ham, sausages, meat pies, kebabs, burgers and nuggets. For each of the above variables, participants were asked about their frequency of consumption through selecting one of the following: never, less than once a week, once a week, 2–4 times a week, 5–6 times a week, once or more daily. For the above variables, frequency categories were converted to a continuous weekly score, as follows: a value of zero was allocated to ‘never’, 0.5 for ‘less than once a week’, one for ‘once a week’, three for ‘2–4 times a week’, 5.5 for ‘5–6 times a week’ and seven for ‘once or more daily’. Salad/raw vegetables (1299) and cooked vegetables (1289) variables were also reported as tablespoons/day. In addition, we used the following categorical diet variables: milk type (1418), which was converted to animal milk (full-cream, semi-skimmed, skimmed), plant milk (soya), never/rarely have milk, and other (other type of milk); spread type (1428), which was converted to animal based spread (butter/spreadable butter), plant based spread (flora pro-active/Benecol), never/rarely use spread, and other (other type of spread/margarine). Alcohol intake was also calculated using data on: beer (1588), alcopop (5364), wine (1568 and 1578), fortified wine (1608) and spirits (1598) consumption in an average week. Each drink was converted into equivalent standard units using the guidelines from the Office for National Statistics to convert volumes to units, where one unit contains 10 ml of ethyl alcohol. Total weekly units of alcohol were calculated by adding the units of beer, wine, and spirits. Excess alcohol consumption was defined as the consumption of more than 14 units of alcohol a week based on the current National Health Service (NHS) guidelines.23 A healthy eating score was also calculated using current guidelines on healthy eating.24 Participants were given a score of one or zero based on their engagement with the current healthy eating guidelines: red meat, ≤3 portions per week = 1 and >3 portions per week = 0; processed meat, ≤1 portion per week = 1 and >1 portion per week = 0; fruit and vegetable consumption, ≥5 per day = 1 and <5 per day = 0; oily fish, ≥1 per week = 1 and <1 per week = 0; non-oily fish, ≥2 per week = 1 and <2 per week = 0; alcohol intake, never drink/≤14 units per week = 1 and >14 units/week = 0. Higher scores therefore indicated healthier eating. Measures of physical activity and physical function Measures of physical activity included: number of days/week walked 10 or more minutes (864); number of days/week of moderate-intensity physical activity undertaken for 10 or more minutes (884); and number of days/week of vigorous-intensity physical activity undertaken for 10 or more minutes (904). The UK Biobank self-reported touchscreen questionnaire was used to capture usual walking pace (924) at baseline; participants were also asked to answer the following question: ‘How would you describe your usual walking pace? (a) slow; (b) steady/average; and (c) brisk’. Objectively measured handgrip strength was also assessed as described elsewhere.6 Other variables Data were also captured for other variables including age (21003; years); systolic blood pressure (BP) (4080; mm Hg); smoking status (20116; current, previous, never, prefer not to answer); total cholesterol (30690; mmol/l) and HDL cholesterol (30760; mmol/l). Total cholesterol and HDL cholesterol serum samples were measured using the Beckman Coulter AU5800 analytical platform by enzymatic and enzyme immuno-inhibition, respectively. Outcomes The study outcomes were all-cause and cardiovascular mortality. Date and cause of death were obtained from the NHS Information Centre for participants from England and Wales, and from the NHS Central Register for participants from Scotland. Cause of death was classified using DF 40001 and the International Classification of Diseases (ICD-10) code assigned to the underlying (primary) cause of death. We defined cardiovascular mortality using ICD-10 codes I00–I79. Cohort definition From the initial sample of 502,621 participants, we excluded people with prevalent cancer (number of self–reported cancers (DF 134); n = 41,706) or prevalent CVD (defined as peripheral vascular disease, angina, heart attack/myocardial infarction, heart failure/pulmonary oedema, stroke and transient ischaemic attack (DF 20002); n = 28,078). From the remaining sample, subjects with information available for all covariates (n = 298,829) were included (Figure 1). Patients were followed-up between study entry until date of death or censoring dates (31 January 2016 for England and Wales; 30 November 2015 for Scotland). Figure 1. Open in new tabDownload slide Participantflowchart. Statistical analysis This study used the concordance index (Harrell’s C-index) to quantify model predictive discrimination ability. The C-index has been frequently used as a metric for evaluating the performance of prognostic models: in survival analysis, C-index is the probability of concordance between the predicted and observed survival, with values ranging from 0.5 (no discrimination) to 1.0 (perfect discrimination).25 The added prognostic value of each investigated factor was quantified as the difference (Δ) in C-index compared with the base model. To address the key aims of the paper, two prognostic models were used. In Model 1, the dietary, physical activity and physical function factors used in this study were added individually to a sex-stratified base model that was adjusted for age. In order to assess the comparative relevance of the resulting C-index statistics, each factor from the SCORE model (smoking status, systolic blood pressure, total and HDL cholesterol) was also added individually to the age-adjusted model to allow for a comparison between investigated and established risk factors. In Model 2, dietary, physical activity and physical function factors were added individually to a sex-stratified base model containing all the SCORE risk factors to assess whether each investigated factor provided added prognostic discrimination when added on top of well-established risk factors. We complemented C-index metric with adjusted R2 estimation (i.e. the explained variation of the survival outcome) in models with and without each investigated factor.26 Finally, in order to highlight differences between prognostic outcomes and those typically reported for observational research, we also report the sex-specific association of each variable with all-cause and cardiovascular mortality using Cox proportional hazard models with time to event measured from study entry (baseline) to death in model 1 (age only) and model 2 (SCORE risk factors: age, smoking status, systolic blood pressure, total and HDL cholesterol). All analyses were conducted using Stata MP 14.1 (Stata Corporation, College Station, Texas, USA). Statistical significance was set at a conservative level of p < 0.01 to account for multiple testing and results reported with a 99% confidence interval (CI). Results Supplementary Material Tables S1 and S2 show the characteristics of the 298,829 participants included in this study. The median age of participants was 57 (interquartile range (IQR), 49–62) years and 159,832 (53.5%) were women. There were 2174 (0.72%) and 3522 (1.17%) all–cause and 286 (0.10%) and 796 (0.27%) CVD deaths in women and men, respectively, during a median (IQR) of 6.92 (6.25–7.56) years of follow-up (2,061,373 person-years). For all-cause mortality, when added to model 1 (age), smoking status and walking pace provided the greatest risk discrimination in both women and men (Supplementary Material Figure S3 and Table 1; Supplementary Material Table S3). When adding diet, physical activity and physical function factors to model 2 (SCORE risk factors), in women the C-index for all-cause mortality was only improved by walking pace, while in men it was improved by walking pace, handgrip strength, cereal intake, healthy eating score, and processed meat intake (Supplementary Material Figure S3 and Table 1; Supplementary Material Table S4). Table 1. C-index significant variables for all-cause and cardiovascular mortality (models 1 and 2). Outcome . Model . Sex . Variable . C-index (99% CI) . C-index difference (99% CI) . All-cause mortality Model 1 Women Smoking status 0.695 (0.680–0.709) 0.020 (0.013–0.028) Usual walking pace 0.687 (0.673–0.702) 0.013 (0.007–0.020) Oily fish intake (portions per week) 0.675 (0.660–0.690) 0.001 (0.000–0.001) Men Smoking status 0.706 (0.695–0.717) 0.025 (0.019–0.031) Usual walking pace 0.704 (0.692–0.715) 0.022 (0.017–0.028) Hand grip strength 0.689 (0.678–0.700) 0.008 (0.004–0.012) Cereal intake (bowls per week) 0.687 (0.676–0.698) 0.006 (0.003–0.009) Processed meat intake (portions per week) 0.685 (0.673–0.696) 0.003 (0.001–0.006) Healthy eating score 0.684 (0.673–0.696) 0.003 (0.001–0.006) Total cholesterol 0.684 (0.673–0.695) 0.003 (0.000–0.005) Fresh fruit intake (pieces per day) 0.683 (0.672–0.695) 0.002 (0.000–0.004) Dried fruit intake (pieces per day) 0.683 (0.672–0.695) 0.002 (0.000–0.004) Beef intake (servings per week) 0.683 (0.672–0.695) 0.002 (0.000–0.004) Milk type 0.683 (0.672–0.694) 0.002 (0.000–0.003) Spread type 0.683 (0.671–0.694) 0.002 (0.000–0.003) Model 2 Women Usual walking pace 0.705 (0.690–0.720) 0.008 (0.003–0.012) Men Usual walking pace 0.722 (0.711–0.733) 0.013 (0.009–0.017) Hand grip strength 0.716 (0.705–0.727) 0.007 (0.004–0.010) Cereal intake (bowls per day) 0.711 (0.700–0.722) 0.002 (0.000–0.004) Healthy eating score 0.711 (0.700–0.722) 0.002 (0.000–0.003) Processed meat intake (portions per week) 0.711 (0.700–0.722) 0.002 (0.000–0.003) Cardiovascular mortality Model 1 Women Smoking status 0.745 (0.707–0.783) 0.030 (0.009–0.052) Usual walking pace 0.738 (0.699–0.778) 0.023 (0.005–0.042) Men Usual walking pace 0.709 (0.685–0.733) 0.034 (0.019–0.048) Smoking status 0.698 (0.674–0.722) 0.022 (0.010–0.035) Cereal intake (bowls per day) 0.684 (0.660–0.707) 0.008 (0.001–0.015) Healthy eating score 0.682 (0.658–0.707) 0.007 (0.000–0.014) Model 2 Women – – – Men Usual walking pace 0.727 (0.704–0.751) 0.024 (0.012–0.035) Hand grip strength 0.712 (0.688–0.736) 0.008 (0.001–0.016) Cereal intake (bowls per day) 0.708 (0.685–0.731) 0.004 (0.000–0.008) Outcome . Model . Sex . Variable . C-index (99% CI) . C-index difference (99% CI) . All-cause mortality Model 1 Women Smoking status 0.695 (0.680–0.709) 0.020 (0.013–0.028) Usual walking pace 0.687 (0.673–0.702) 0.013 (0.007–0.020) Oily fish intake (portions per week) 0.675 (0.660–0.690) 0.001 (0.000–0.001) Men Smoking status 0.706 (0.695–0.717) 0.025 (0.019–0.031) Usual walking pace 0.704 (0.692–0.715) 0.022 (0.017–0.028) Hand grip strength 0.689 (0.678–0.700) 0.008 (0.004–0.012) Cereal intake (bowls per week) 0.687 (0.676–0.698) 0.006 (0.003–0.009) Processed meat intake (portions per week) 0.685 (0.673–0.696) 0.003 (0.001–0.006) Healthy eating score 0.684 (0.673–0.696) 0.003 (0.001–0.006) Total cholesterol 0.684 (0.673–0.695) 0.003 (0.000–0.005) Fresh fruit intake (pieces per day) 0.683 (0.672–0.695) 0.002 (0.000–0.004) Dried fruit intake (pieces per day) 0.683 (0.672–0.695) 0.002 (0.000–0.004) Beef intake (servings per week) 0.683 (0.672–0.695) 0.002 (0.000–0.004) Milk type 0.683 (0.672–0.694) 0.002 (0.000–0.003) Spread type 0.683 (0.671–0.694) 0.002 (0.000–0.003) Model 2 Women Usual walking pace 0.705 (0.690–0.720) 0.008 (0.003–0.012) Men Usual walking pace 0.722 (0.711–0.733) 0.013 (0.009–0.017) Hand grip strength 0.716 (0.705–0.727) 0.007 (0.004–0.010) Cereal intake (bowls per day) 0.711 (0.700–0.722) 0.002 (0.000–0.004) Healthy eating score 0.711 (0.700–0.722) 0.002 (0.000–0.003) Processed meat intake (portions per week) 0.711 (0.700–0.722) 0.002 (0.000–0.003) Cardiovascular mortality Model 1 Women Smoking status 0.745 (0.707–0.783) 0.030 (0.009–0.052) Usual walking pace 0.738 (0.699–0.778) 0.023 (0.005–0.042) Men Usual walking pace 0.709 (0.685–0.733) 0.034 (0.019–0.048) Smoking status 0.698 (0.674–0.722) 0.022 (0.010–0.035) Cereal intake (bowls per day) 0.684 (0.660–0.707) 0.008 (0.001–0.015) Healthy eating score 0.682 (0.658–0.707) 0.007 (0.000–0.014) Model 2 Women – – – Men Usual walking pace 0.727 (0.704–0.751) 0.024 (0.012–0.035) Hand grip strength 0.712 (0.688–0.736) 0.008 (0.001–0.016) Cereal intake (bowls per day) 0.708 (0.685–0.731) 0.004 (0.000–0.008) CI: confidence interval. * Shown are variables with significant (p < 0.01) change in C-index compared with base models (all C-index values are reported in Table S2 and S3). Within each model and sex, variables are sorted by descending order of C-index difference. All-cause mortality: †Reference C-index (99% CI) for base model 1 (age adjusted): Women: 0.674 (0.663–0.685); Men: 0.681 (0.673–0.690) All-cause mortality: ‡Reference C-index (99% CI) for base model 2 (SCORE variables): Women: 0.697 (0.686–0.709); Men: 0.709 (0.701–0.718) Cardiovascular mortality: †Reference C-index (99% CI) for base model 1 (age adjusted): Women: 0.715 (0.685–0.744); Men: 0.676 (0.657–0.694) Cardiovascular mortality: ‡Reference C-index (99% CI) for base model 2 (SCORE variables): Women: 0.756 (0.729–0.783); Men: 0.704 (0.686–0.722) ‘-’ indicates non-significant. Open in new tab Table 1. C-index significant variables for all-cause and cardiovascular mortality (models 1 and 2). Outcome . Model . Sex . Variable . C-index (99% CI) . C-index difference (99% CI) . All-cause mortality Model 1 Women Smoking status 0.695 (0.680–0.709) 0.020 (0.013–0.028) Usual walking pace 0.687 (0.673–0.702) 0.013 (0.007–0.020) Oily fish intake (portions per week) 0.675 (0.660–0.690) 0.001 (0.000–0.001) Men Smoking status 0.706 (0.695–0.717) 0.025 (0.019–0.031) Usual walking pace 0.704 (0.692–0.715) 0.022 (0.017–0.028) Hand grip strength 0.689 (0.678–0.700) 0.008 (0.004–0.012) Cereal intake (bowls per week) 0.687 (0.676–0.698) 0.006 (0.003–0.009) Processed meat intake (portions per week) 0.685 (0.673–0.696) 0.003 (0.001–0.006) Healthy eating score 0.684 (0.673–0.696) 0.003 (0.001–0.006) Total cholesterol 0.684 (0.673–0.695) 0.003 (0.000–0.005) Fresh fruit intake (pieces per day) 0.683 (0.672–0.695) 0.002 (0.000–0.004) Dried fruit intake (pieces per day) 0.683 (0.672–0.695) 0.002 (0.000–0.004) Beef intake (servings per week) 0.683 (0.672–0.695) 0.002 (0.000–0.004) Milk type 0.683 (0.672–0.694) 0.002 (0.000–0.003) Spread type 0.683 (0.671–0.694) 0.002 (0.000–0.003) Model 2 Women Usual walking pace 0.705 (0.690–0.720) 0.008 (0.003–0.012) Men Usual walking pace 0.722 (0.711–0.733) 0.013 (0.009–0.017) Hand grip strength 0.716 (0.705–0.727) 0.007 (0.004–0.010) Cereal intake (bowls per day) 0.711 (0.700–0.722) 0.002 (0.000–0.004) Healthy eating score 0.711 (0.700–0.722) 0.002 (0.000–0.003) Processed meat intake (portions per week) 0.711 (0.700–0.722) 0.002 (0.000–0.003) Cardiovascular mortality Model 1 Women Smoking status 0.745 (0.707–0.783) 0.030 (0.009–0.052) Usual walking pace 0.738 (0.699–0.778) 0.023 (0.005–0.042) Men Usual walking pace 0.709 (0.685–0.733) 0.034 (0.019–0.048) Smoking status 0.698 (0.674–0.722) 0.022 (0.010–0.035) Cereal intake (bowls per day) 0.684 (0.660–0.707) 0.008 (0.001–0.015) Healthy eating score 0.682 (0.658–0.707) 0.007 (0.000–0.014) Model 2 Women – – – Men Usual walking pace 0.727 (0.704–0.751) 0.024 (0.012–0.035) Hand grip strength 0.712 (0.688–0.736) 0.008 (0.001–0.016) Cereal intake (bowls per day) 0.708 (0.685–0.731) 0.004 (0.000–0.008) Outcome . Model . Sex . Variable . C-index (99% CI) . C-index difference (99% CI) . All-cause mortality Model 1 Women Smoking status 0.695 (0.680–0.709) 0.020 (0.013–0.028) Usual walking pace 0.687 (0.673–0.702) 0.013 (0.007–0.020) Oily fish intake (portions per week) 0.675 (0.660–0.690) 0.001 (0.000–0.001) Men Smoking status 0.706 (0.695–0.717) 0.025 (0.019–0.031) Usual walking pace 0.704 (0.692–0.715) 0.022 (0.017–0.028) Hand grip strength 0.689 (0.678–0.700) 0.008 (0.004–0.012) Cereal intake (bowls per week) 0.687 (0.676–0.698) 0.006 (0.003–0.009) Processed meat intake (portions per week) 0.685 (0.673–0.696) 0.003 (0.001–0.006) Healthy eating score 0.684 (0.673–0.696) 0.003 (0.001–0.006) Total cholesterol 0.684 (0.673–0.695) 0.003 (0.000–0.005) Fresh fruit intake (pieces per day) 0.683 (0.672–0.695) 0.002 (0.000–0.004) Dried fruit intake (pieces per day) 0.683 (0.672–0.695) 0.002 (0.000–0.004) Beef intake (servings per week) 0.683 (0.672–0.695) 0.002 (0.000–0.004) Milk type 0.683 (0.672–0.694) 0.002 (0.000–0.003) Spread type 0.683 (0.671–0.694) 0.002 (0.000–0.003) Model 2 Women Usual walking pace 0.705 (0.690–0.720) 0.008 (0.003–0.012) Men Usual walking pace 0.722 (0.711–0.733) 0.013 (0.009–0.017) Hand grip strength 0.716 (0.705–0.727) 0.007 (0.004–0.010) Cereal intake (bowls per day) 0.711 (0.700–0.722) 0.002 (0.000–0.004) Healthy eating score 0.711 (0.700–0.722) 0.002 (0.000–0.003) Processed meat intake (portions per week) 0.711 (0.700–0.722) 0.002 (0.000–0.003) Cardiovascular mortality Model 1 Women Smoking status 0.745 (0.707–0.783) 0.030 (0.009–0.052) Usual walking pace 0.738 (0.699–0.778) 0.023 (0.005–0.042) Men Usual walking pace 0.709 (0.685–0.733) 0.034 (0.019–0.048) Smoking status 0.698 (0.674–0.722) 0.022 (0.010–0.035) Cereal intake (bowls per day) 0.684 (0.660–0.707) 0.008 (0.001–0.015) Healthy eating score 0.682 (0.658–0.707) 0.007 (0.000–0.014) Model 2 Women – – – Men Usual walking pace 0.727 (0.704–0.751) 0.024 (0.012–0.035) Hand grip strength 0.712 (0.688–0.736) 0.008 (0.001–0.016) Cereal intake (bowls per day) 0.708 (0.685–0.731) 0.004 (0.000–0.008) CI: confidence interval. * Shown are variables with significant (p < 0.01) change in C-index compared with base models (all C-index values are reported in Table S2 and S3). Within each model and sex, variables are sorted by descending order of C-index difference. All-cause mortality: †Reference C-index (99% CI) for base model 1 (age adjusted): Women: 0.674 (0.663–0.685); Men: 0.681 (0.673–0.690) All-cause mortality: ‡Reference C-index (99% CI) for base model 2 (SCORE variables): Women: 0.697 (0.686–0.709); Men: 0.709 (0.701–0.718) Cardiovascular mortality: †Reference C-index (99% CI) for base model 1 (age adjusted): Women: 0.715 (0.685–0.744); Men: 0.676 (0.657–0.694) Cardiovascular mortality: ‡Reference C-index (99% CI) for base model 2 (SCORE variables): Women: 0.756 (0.729–0.783); Men: 0.704 (0.686–0.722) ‘-’ indicates non-significant. Open in new tab For cardiovascular mortality, when added to model 1 smoking status provided the greatest discrimination in women, whilst walking pace did in men (Supplementary Material Figure S3 and Table 1; Supplementary Material Table S3). In model 2, none of the lifestyle factors improved the C-index for cardiovascular mortality in women; conversely, in men it was improved by walking pace, handgrip strength, and cereal intake (Supplementary Material Figure S3 and Table 1; Supplementary Material Table S4). The pattern of results for R2 mirrored those of the C-index, for both all-cause and cardiovascular mortality (Supplementary Material Figure S4). Model 2 R2 for all-cause mortality was 0.294 and 0.323 in women and men, respectively; these values increased to 0.326 and 0.374, respectively, upon the inclusion of walking pace; corresponding estimates for cardiovascular mortality were 0.460 and 0.319 for model 2 and 0.515 and 0.405 upon the inclusion of walking pace (Supplementary Material Figure S4). More associations were observed when quantified by hazard ratios compared with the prognostic factors identified with the C-index. In age-adjusted models, all-cause mortality hazard ratios were significant for nine and 13 dietary factors, and for five and three physical activity and physical function factors, in women and men, respectively; corresponding values for cardiovascular mortality were two and seven (dietary factors) and one and three (physical activity and function factors) (Supplementary Material Figure S5). For the SCORE risk factors, total cholesterol and HDL cholesterol were negatively associated with all-cause mortality in women and men, while systolic blood pressure was positively associated with all-cause mortality in men. HDL cholesterol was negatively associated with cardiovascular mortality in women and systolic blood pressure was positively associated with cardiovascular mortality in men. All-cause and cardiovascular mortality hazard ratios estimates for each lifestyle when adjusted for the SCORE variables are displayed in Supplementary Material Figure S6. Discussion The aim of this study was to investigate the prognostic relevance of simple measures of diet, physical activity and physical function in comparison with, and when added to, established risk factors used to predict cardiovascular mortality. Several factors were highlighted as improving prediction, the most notable of which was walking pace, which improved predictive discrimination for of all-cause and cardiovascular mortality to a similar extent as smoking status in age-adjusted models, particularly for men. Importantly, walking pace also improved predictive discrimination for both all-cause mortality (women and men) and cardiovascular mortality (men) when added to a base model containing the conventional SCORE12 risk factors age, smoking status, systolic blood pressure, total and HDL cholesterol, which are also widely used in other established risk prediction models, including Pooled Cohort Equations (PCE),15 Framingham Risk Score (FRS)14 and Reynolds Risk Score (RRS).13 Handgrip strength was also found to improve discrimination for both all-cause mortality and cardiovascular mortality in men, but to a lesser extent compared with walking pace. These results highlight that a simple measure of self-reported walking pace may be an important prognostic marker of all-cause and cardiovascular mortality when added to established risk factors commonly used in risk prediction tools. There is wealth of literature investigating the association between lifestyle factors and mortality outcomes, with diet and physical activity thought to directly influence the risk of all-cause and cardiovascular mortality, including previous work conducted by our group.2–4,27–30 However, few studies have systematically investigated diet and physical activity in relation to mortality prediction, with previous research restricted to assessing the predictive discrimination of lifestyle scores incorporating a limited number of factors.9,10 As far as we are aware, this is the first study to systematically compare the predictive discrimination of multiple dietary, physical activity and physical function measures with the aim of highlighting those with the greatest predictive potential. The finding that walking pace provides a similar magnitude of improvement in predictive discrimination as smoking for all-cause and cardiovascular mortality outcomes and that it adds predictive information when added to established risk factors highlights it as a simple, non-invasive measure with potential to improve the performance of established risk prediction tools. These findings add novel prognostic information to previous, more aetiological research which demonstrated a strong association between walking pace and health outcomes. For example, slow walkers have been shown to have between 2–4 times higher risk of all-cause and cardiovascular mortality compared with fast walkers, as evidenced by hazard ratio values in the present and previous UK Biobank publications.6,31 Recent research has also found that fast walkers have a longer life expectancy across all categories of normal weight and obesity status.32 Self-reported walking pace has also been shown to be more strongly associated with mortality outcomes than measures of physical activity volume,31,33 which further corresponds to the findings of this study where the frequency of walking or other physical activity did not improve prediction of all-cause or cardiovascular mortality. Walking pace is a widely used measure of functional status and physical frailty, and self-reported walking pace has been shown to be strongly associated with cardiorespiratory fitness,6,34 which is considered a clinical vital sign.35 A slow self-reported walking pace is therefore likely to identify people with low physiological reserve and an impaired ability to resist physical and/or psychological stressors. Our findings suggest that the consistent evidence of an association between walking pace and health outcomes translates also into improved risk prediction. Out of all the dietary variables considered, when added to established risk factors none improved risk discrimination in women whilst only cereal intake improved discrimination for both all-cause and cardiovascular mortality in men, despite multiple dietary factors displaying an association with mortality outcomes when assessed through hazard ratios in both men and women (i.e. red and processed meat). The discrepancy in findings between association (measured using hazard ratio) and prediction (measured using the C-index) for dietary factors further highlights that establishing an association does not necessarily translate into improved prediction. The importance of cereal intake in the association with, and prediction of, mortality outcomes observed in this study may be related to fibre intake. A recent systematic review and meta-analysis of prospective cohort studies suggests that cereal fibre intake is inversely associated with all-cause and cardiovascular mortality.36 Our study adds to this evidence by further suggesting that cereal intake could improve risk prediction in men. There are some strengths and limitations that deserve to be mentioned. Strengths include the large contemporary sample of participants, the follow-up duration allowing a large number of events, and the availability of multiple risk factors, including cholesterol, blood pressure and smoking. Nevertheless, there are a number limitations. There were more deaths in men than women, with a higher proportion dying of CVD; therefore, the lower number of lifestyle factors found to improve prediction in women compared with men may be related to a lower number of events.25 Diet was assessed using a food frequency questionnaire, therefore the measure of dietary variables included in this analysis may have some degree of measurement error whilst limiting the ability to capture energy intake and macronutrient content. However, the food frequency questionnaire used in UK Biobank has been shown to have reasonable reliability for the dietary variables used in this study.37 Physical activity, physical function and dietary intake were self-reported, thus residual confounding due to limited measurement precision also remains a possibility. The limitations around confounding and causality in observational research are, however, less relevant for prognostic research, where the aim is to investigate whether factors aid risk prediction, as opposed to interest in identifying factors that can be used for health promotion (aetiological research). UK Biobank is not representative of the general population; this may limit the generalizability of these findings to other populations.38 In conclusion, this study suggests that amongst several easily measured dietary, physical activity and physical function measures, self-reported walking pace has the potential to significantly improve the prediction of all-cause and cardiovascular mortality beyond established risk factors. Future studies in other populations are needed to validate the role of walking pace on top of established risk assessment tools. Acknowledgements The authors would like to thank the participants of the UK Biobank. This research has been conducted using the UK Biobank Resource under Application Number 3140. Author contribution SA contributed to conception and design, data acquisition, analysis and interpretation and drafted the manuscript. FZ and TY designed the study, contributed to data acquisition, interpretation and analysis and critically revised the manuscript. MJD and KK contributed to acquisition and interpretation of the study and critically revised the manuscript. All authors gave final approval for the manuscript and agreed to be accountable for all aspects of this work ensuring its 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: MJD: consultant, advisory board member and speaker/fees and grants in support of investigator/investigator-initiated trial from Novo Nordisk, Sanofi-Aventis, Lilly, Boehringer Ingelheim and Janssen. Consultant, advisory board member and speaker of Merck Sharp and Dohme and AstraZeneca. Advisory board member of Servier. Speaker fees from Mitsubishi Tanabe Pharma Corporation and Takeda Pharmaceuticals International Inc. KK: consultant and speaker fees from Amgen, Bayer, Novartis, Roche and Servier. Consultant, advisory board member and speaker fees from AstraZeneca, Novo Nordisk, Sanofi-Aventis, Lilly, Merck, Sharp and Dohme. Grant in support of investigator and investigator-initiated trials from Pfizer, Boehringer Ingelheim. In support of investigator and investigator initiated trials from AstaZeneca, Novartis, Novo Nordisk, Sanofi-Aventis, Lilly, and Merck Sharp and Dohme. TY: grant research supported by the NIHR Leicester BRC. SA and FZ: none. 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 the National Institute for Health Research (NIHR) Leicester Biomedical Research Centre, Leicester, UK to SA, TY and MJD; NIHR Applied Research Collaboration East Midlands (ARC EM) to KK; FZ is a clinical research fellow funded with an unrestricted educational grant from the NIHR ARC East Midlands to the University of Leicester. The funding source had no role in study design, data collection, data analysis, data interpretation or writing of the report. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care. References 1 Nichols M , Townsend N, Scarborough P, et al. Cardiovascular disease in Europe 2014: Epidemiological update . Eur Heart J 2014 ; 35 : 2950 – 2959 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Arem H , Moore SC, Patel A, et al. Leisure time physical activity and mortality: A detailed pooled analysis of the dose-response relationship . JAMA Intern Med 2015 ; 175 : 959 – 967 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Nocon M , Hiemann T, Müller-Riemenschneider F, et al. Association of physical activity with all-cause and cardiovascular mortality: A systematic review and meta-analysis . Eur J Cardiovasc Prev Rehabil 2008 ; 15 : 239 – 246 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Sofi F , Cesari F, Abbate R, et al. Adherence to Mediterranean diet and health status: Meta-analysis . BMJ 2008 ; 337 : a1344 – a1344 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Jacobs DR Jr, Adachi H, et al. Cigarette smoking and mortality risk: Twenty-five-year follow-up of the seven countries study . Arch Intern Med 1999 ; 159 : 733 – 740 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Yates T , Zaccardi F, Dhalwani NN, et al. Association of walking pace and handgrip strength with all-cause, cardiovascular, and cancer mortality: A UK Biobank observational study . Eur Heart J 2017 ; 38 : 3232 – 3240 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Ganna A , Ingelsson E. 5 Year mortality predictors in 498 103 UK Biobank participants: A prospective population-based study . Lancet 2015 ; 386 : 533 – 540 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Kodama S , Saito K, Tanaka S, et al. Cardiorespiratory fitness as a quantitative predictor of all-cause mortality and cardiovascular events in healthy men and women: A meta-analysis . JAMA 2009 ; 301 : 2024 – 2035 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Ding D , Rogers K, van der Ploeg H, et al. Traditional and emerging lifestyle risk behaviors and all-cause mortality in middle-aged and older adults: Evidence from a large population-based Australian cohort . PLoS Med 2015 ; 12 : e1001917 – e1001917 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Walter S , Mackenbach J, Vokó Z, et al. Genetic, physiological, and lifestyle predictors of mortality in the general population . Am J Public Health 2012 ; 102 : e3 – e10 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Hemingway H , Croft P, Perel P, et al. Prognosis research strategy (PROGRESS) 1: A framework for researching clinical outcomes . BMJ 2013 ; 346 : e5595 – e5595 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Conroy RM , Pyörälä K, Fitzgerald AP, et al. Estimation of ten-year risk of fatal cardiovascular disease in Europe: The SCORE project . Eur Heart J 2003 ; 24 : 987 – 1003 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Ridker PM , Buring JE, Rifai N, et al. Development and validation of improved algorithms for the assessment of global cardiovascular risk in women: The Reynolds Risk Score . JAMA 2007 ; 297 : 611 – 619 . Google Scholar Crossref Search ADS PubMed WorldCat 14 D’Agostino RB , Vasan RS, Pencina MJ, et al. General cardiovascular risk profile for use in primary care: The Framingham Heart Study . Circulation 2008 ; 117 : 743 – 753 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Catapano AL , Graham I, De Backer G, et al. 2016 ESC/EAS guidelines for the management of dyslipidaemias . Eur Heart J 2016 ; 37 : 2999 – 3058 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Tomasik T , Krzysztoń J, Dubas-Jakóbczyk K, et al. The systematic coronary risk evaluation (SCORE) for the prevention of cardiovascular diseases. Does evidence exist for its effectiveness? A systematic review . Acta Cardiologica 2017 ; 72 : 370 – 379 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Cuende JI , Cuende N, Calaveras-Lagartos J. How to calculate vascular age with the SCORE project scales: A new method of cardiovascular risk evaluation . Eur Heart J 2010 ; 31 : 2351 – 2358 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Neuhauser HK , Ellert U, Kurth B-M. A comparison of Framingham and SCORE-based cardiovascular risk estimates in participants of the German National Health Interview and Examination Survey 1998 . Eur J Cardiovasc Prev Rehabil 2005 ; 12 : 442 – 450 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Siontis GCM , Tzoulaki I, Siontis KC, et al. Comparisons of established risk prediction models for cardiovascular disease: Systematic review . BMJ 2012 ; 344 : e3318 – e3318 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Silventoinen K , Pankow J, Lindstrom J, et al. The validity of the Finnish Diabetes Risk Score for the prediction of the incidence of coronary heart disease and stroke, and total mortality . Eur J Cardiovasc Prev Rehabil 2005 ; 12 : 451 – 458 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Hlatky MA , Greenland P, Arnett DK, et al. Criteria for evaluation of novel markers of cardiovascular risk: A scientific statement from the american heart association . Circulation 2009 ; 119 : 2408 – 2416 . Google Scholar Crossref Search ADS PubMed WorldCat 22 UK Biobank. Data showcase, http://www.ukbiobank.ac.uk/data-showcase/ (2017; accessed 3 June 2019) . 23 NHS. Alcohol units, https://www.nhs.uk/live-well/alcohol-support/calculating-alcohol-units/ (2018, accessed 3 June 2019) . 24 Public Health England. A quick guide to the government's healthy eating recommendations. London. PHE publications, https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/742746/A_quick_guide_to_govt_healthy_eating_update.pdf (2018, accessed 1 June 2019) . 25 Harrell FE Jr, Califf RM, Pryor DB, et al. Evaluating the yield of medical tests . JAMA 1982 ; 247 : 2543 – 2546 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Royston P . Explained variation for survival models . Stata J 2006 ; 6 : 83 – 96 . Google Scholar Crossref Search ADS WorldCat 27 Sattelmair J , Pertman J, Ding EL, et al. Dose response between physical activity and risk of coronary heart disease: A meta-analysis . Circulation 2011 ; 124 : 789 – 795 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Aune D , Giovannucci E, Boffetta P, et al. Fruit and vegetable intake and the risk of cardiovascular disease, total cancer and all-cause mortality-a systematic review and dose-response meta-analysis of prospective studies . Int J Epidemiol 2017 ; 46 : 1029 – 1056 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Argyridou S , Zaccardi F, Davies MJ, et al. Relevance of physical function in the association of red and processed meat intake with all–cause, cardiovascular, and cancer mortality . Nutr Metab Cardiovasc Dis . Epub ahead of print 27 June 2019. DOI: 10.1016/j.numecd.2019.06.019 . OpenURL Placeholder Text WorldCat 30 Kwok CS , Gulati M, Michos ED, et al. Dietary components and risk of cardiovascular disease and all-cause mortality: A review of evidence from meta-analyses . Eur J Prev Cardiol 2019 ; 26 : 1415 – 1429 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Stamatakis E , Ekelund U, Ding D, et al. Is the time right for quantitative public health guidelines on sitting? A narrative review of sedentary behaviour research paradigms and findings . Brit J Sport Med 2019 ; 53 : 377 – 382 . Google Scholar Crossref Search ADS WorldCat 32 Zaccardi F , Davies MJ, Khunti K, et al. Comparative relevance of physical fitness and adiposity on life expectancy . Mayo Clin Proc 2019 ; 94 : 985 – 994 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Hamer M , Chida Y. Walking and primary prevention: A meta-analysis of prospective cohort studies . Brit J Sport Med 2008 ; 42 : 238 – 243 . Google Scholar Crossref Search ADS WorldCat 34 Fried LP , Tangen CM, Walston J, et al. Frailty in older adults: Evidence for a phenotype . J Gerontol A Biol Sci Med Sci 2001 ; 56 : M146 – M156 . Google Scholar Crossref Search ADS PubMed WorldCat 35 Ross R , Blair SN, Arena R, et al. Importance of assessing cardiorespiratory fitness in clinical practice: A case for fitness as a clinical vital sign: A scientific statement from the American Heart Association . Circulation 2016 ; 134 : e653 – e699 . Google Scholar Crossref Search ADS PubMed WorldCat 36 Hajishafiee M , Saneei P, Benisi-Kohansal S, et al. Cereal fibre intake and risk of mortality from all causes, CVD, cancer and inflammatory diseases: A systematic review and meta-analysis of prospective cohort studies . Br J Nutr 2016 ; 116 : 343 – 352 . Google Scholar Crossref Search ADS PubMed WorldCat 37 Bradbury KE , Young HJ, Guo W, et al. Dietary assessment in UK Biobank: An evaluation of the performance of the touchscreen dietary questionnaire . J Nutr Sci 2018 ; 7 : e6 – e6 . Google Scholar Crossref Search ADS PubMed WorldCat 38 Fry A , Littlejohns TJ, Sudlow C, et al. Comparison of sociodemographic and health-related characteristics of UK Biobank participants with those of the general population . Am J Epidemiol 2017 ; 186 : 1026 – 1034 . Google Scholar Crossref Search ADS PubMed WorldCat © The European Society of Cardiology 2020 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 2020

Oldridge, Neil; Taylor, Rod S

doi: 10.1177/2047487319881839pmid: 31657233

Aims Prescribed exercise is effective in adults with coronary heart disease (CHD), chronic heart failure (CHF), intermittent claudication, body mass index (BMI) ≥25 kg/m2, hypertension or type 2 diabetes mellitus (T2DM), but the evidence for its cost-effectiveness is limited, shows large variations and is partly contradictory. Using World Health Organization and American Heart Association/American College of Cardiology value for money thresholds, we report the cost-effectiveness of exercise therapy, exercise training and exercise-based cardiac rehabilitation. Methods Electronic databases were searched for incremental cost-effectiveness and incremental cost–utility ratios and/or the probability of cost-effectiveness of exercise prescribed as therapy in economic evaluations conducted alongside randomized controlled trials (RCTs) published between 1 July 2008 and 28 October 2018. Results Of 19 incremental cost–utility ratios reported in 15 RCTs in patients with CHD, CHF, intermittent claudication or BMI ≥25 kg/m2, 63% met both value for money thresholds as ‘highly cost-effective’ or ‘high value’, with 26% ‘not cost-effective’ or of ‘low value’. The probability of intervention cost-effectiveness ranged from 23 to 100%, probably due to the different populations, interventions and comparators reported in the individual RCTs. Confirmation with the Consolidated Health Economic Evaluation Reporting checklist varied widely across the included studies. Conclusions The findings of this review support the cost-effectiveness of exercise therapy in patients with CHD, CHF, BMI ≥25 kg/m2 or intermittent claudication, but, with concerns about reporting standards, need further confirmation. No eligible economic evaluation based on RCTs was identified in patients with hypertension or T2DM. Cost-effectiveness, systematic reviews, exercise, cardiac rehabilitation, coronary heart disease, chronic heart failure, type 2 diabetes mellitus, hypertension, body mass index, intermittent claudication Introduction Although physical inactivity is associated with a number of chronic diseases,1 the ‘economic burden of physical inactivity remains an important yet underdeveloped area’2 and ‘is a costly pandemic that is associated with a substantial disease burden in almost every country where estimates exist’.2 For example, the annual total direct healthcare expenditure associated with physical inactivity in 142 countries is Int$53.8 billion, with Int$10.5 billion in Europe and Int$24.7 billion in the USA, i.e. 0.64, 0.55% and 0.85%, respectively.3 At the same time, there is strong evidence for the effectiveness of prescribed exercise in patients with chronic disease, specifically coronary heart disease (CHD), chronic heart failure (CHF), intermittent claudication and in patients with body mass index (BMI) ≥25 kg/m2 (overweight or obese), hypertension or type 2 diabetes mellitus (T2DM).4,5 Systematic reviews of evidence published prior to June 2008 on the cost-effectiveness of prescribed physical exercise in patients with chronic diseases is ‘limited’, ‘shows large variation’ and ‘was partly contradictory’,6 whereas evidence published between 2008 and 2016 suggests that ‘exercise programs for the treatment of musculoskeletal and rheumatologic disorders and, to a lesser degree for the treatment of cardiovascular diseases, appear cost-effective’.7 In a third systematic review, evidence published between 2001 and 2016 suggests that cardiac rehabilitation ‘is cost-effective, especially with exercise as a component’.8 Healthcare economic evaluations, defined as the ‘comparative analysis of alternative courses of action in terms of both their costs and consequences’,9 provide clinicians, patients, healthcare planners and policy-makers with important information about the available alternatives in the context of limited resources. Cost-effectiveness ‘value for money’ thresholds for health benefits are generally identified as ‘very good’, ‘relatively good’ or ‘not good’.9 The World Health Organization (WHO) value for money threshold10 is the most commonly cited cost-effectiveness threshold guideline, even though it is often criticized for having major limitations.11 A more recent guideline has been recommended by the American College of Cardiology (ACC) and the American Heart Association (AHA).12 With scarce financial resources for healthcare and high healthcare costs, the primary purpose of this systematic review is to report the cost-effectiveness and the value for money of prescribed exercise therapy, exercise training and exercise-based cardiac rehabilitation in adults with CHD, CHF, intermittent claudication, BMI ≥25 kg/m2, hypertension and T2DM based on randomized controlled trial (RCT) data published between 1 July 2008 and 28 October 2018. Methods Data sources and search strategy The Preferred Reporting for Systematic Reviews and Meta-Analyses13 and the Cochrane Handbook for Interventional Reviews14 guidelines were followed when searching the following electronic databases from inception to 28 October 2018 with no language limit: the Cochrane Controlled Trial Register; the Cumulative Index to Nursing and Allied Health Literature; EMBASE; EMCARE; MEDLINE (OVID); PsycINFO; and PubMed (excluding MEDLINE). The search strategies, developed by an information specialist, with the Ovid MEDLINE search strategy provided in the Supplementary Materials (Table e1), utilized the PICO[S] framework without reference to a specific Comparator term: Population – CHD, CHF, intermittent claudication, BMI ≥25 kg/m2, hypertension, T2DM; Intervention – exercise therapy, exercise training, exercise-based cardiac rehabilitation; Outcomes – costs, life-years, quality-adjusted life-year (QALY), incremental cost-effectiveness ratio (ICER), incremental cost–utility ratio (ICUR) with or without a probability of cost-effectiveness; and Study design – full economic evaluation alongside an RCT. All papers published before 1 July 2008 (the end date for the inclusion of studies in the systematic review of Roine et al.6) were deleted. The remaining titles, keywords and abstracts of each paper published by 28 October 2018 were electronically screened with duplicate titles and the studies focused on pregnancy, children and adolescents, non-RCTs and without a full economic evaluation were deleted. The remaining titles were retrieved for full text screening. The Consolidated Health Economic Evaluation Reporting Standards (CHEERS)15 were used to assess the reporting quality of the economic evaluations with cost-effectiveness analysis of physical activity carried out alongside RCTs. Cost-effectiveness Economic evaluations of prescribed exercise conducted alongside an RCT with an ICER or ICUR, with or without the probability of being cost-effective, were eligible for analysis. When not reported in a paper, the ICUR was calculated from the reported cost and QALY data. The WHO10 and the ACC/AHA12 guidelines were used to determine the value for money for each intervention. The WHO guideline recognizes a ‘highly cost-effective’ threshold at < 1-times national per capita gross domestic product (GDP), ‘cost-effective’ at 1–3-times the GDP and ‘not cost-effective’ at >3-times the GDP,10 with the national GDP figures given in the Supplementary Materials (Table e2). The ACC/AHA guidelines identify < US$50,000/QALY as the high value threshold, US$50,000–150,000/QALY as the intermediate value and >US$150,000/QALY as the low value threshold.12 Results Selection and inclusion of papers A total of 5067 papers with two additional papers in patients with CHD from other sources were identified. Of these, 44 papers with a full economic evaluation alongside an RCT were eligible for full text screening (Table 1). Of these, 15 RCTs with an ICER, an ICUR and/or a probability of cost-effectiveness were identified, eight in patients with CHD,16–23 one in patients with CHF,24 four in patients with intermittent claudication,25–28 and two in patients with BMI ≥25 kg/m2,29,30 with no eligible RCT identified in patients with either hypertension or T2DM. The quality of reporting in the economic evaluations of physical activity considered to be eligible for inclusion in this systematic review of physical activity RCTs is detailed in the Supplementary Material (Figure e1). The quality of reporting can be summarized as the proportion of studies meeting the standards on a particular item as ranging from 20 to 100%, with each of the 15 RCTs meeting the CHEERS checklist standards15 on 54% of the 24 items. Table 1. Summary of results of electronic database searches for exercise therapy, exercise training or exercise-based cardiac rehabilitation economic evaluation randomized controlled trial titles. . CHD . CHF . IC . BMI ≥25 kg/m2 . HTN . T2DM . Total no. of papers identified 975 + 2 555 164 1405 826 1142 Titles excluded (title, keywords, abstract screened) Duplicates 470 241 80 579 353 576 Not RCT 308 191 58 422 286 305 Not exercise 119 73 16 271 139 182 Not eco eval 66 40 6 127 47 74 Full text screened for incremental cost-effectiveness or incremental cost–utility ratio Eco eval RCTs 14 10 8 6 1 5 Eligible exercise eco eval RCTs 8 1 4 2 0 0 . CHD . CHF . IC . BMI ≥25 kg/m2 . HTN . T2DM . Total no. of papers identified 975 + 2 555 164 1405 826 1142 Titles excluded (title, keywords, abstract screened) Duplicates 470 241 80 579 353 576 Not RCT 308 191 58 422 286 305 Not exercise 119 73 16 271 139 182 Not eco eval 66 40 6 127 47 74 Full text screened for incremental cost-effectiveness or incremental cost–utility ratio Eco eval RCTs 14 10 8 6 1 5 Eligible exercise eco eval RCTs 8 1 4 2 0 0 BMI: body mass index >25 kg/m2 (overweight and obese); CHD: coronary heart disease; CHF: chronic heart failure; Eco eval: economic evaluation; HTN: hypertension; IC: intermittent claudication; RCT: randomized controlled trial; T2DM: type 2 diabetes mellitus. Open in new tab Table 1. Summary of results of electronic database searches for exercise therapy, exercise training or exercise-based cardiac rehabilitation economic evaluation randomized controlled trial titles. . CHD . CHF . IC . BMI ≥25 kg/m2 . HTN . T2DM . Total no. of papers identified 975 + 2 555 164 1405 826 1142 Titles excluded (title, keywords, abstract screened) Duplicates 470 241 80 579 353 576 Not RCT 308 191 58 422 286 305 Not exercise 119 73 16 271 139 182 Not eco eval 66 40 6 127 47 74 Full text screened for incremental cost-effectiveness or incremental cost–utility ratio Eco eval RCTs 14 10 8 6 1 5 Eligible exercise eco eval RCTs 8 1 4 2 0 0 . CHD . CHF . IC . BMI ≥25 kg/m2 . HTN . T2DM . Total no. of papers identified 975 + 2 555 164 1405 826 1142 Titles excluded (title, keywords, abstract screened) Duplicates 470 241 80 579 353 576 Not RCT 308 191 58 422 286 305 Not exercise 119 73 16 271 139 182 Not eco eval 66 40 6 127 47 74 Full text screened for incremental cost-effectiveness or incremental cost–utility ratio Eco eval RCTs 14 10 8 6 1 5 Eligible exercise eco eval RCTs 8 1 4 2 0 0 BMI: body mass index >25 kg/m2 (overweight and obese); CHD: coronary heart disease; CHF: chronic heart failure; Eco eval: economic evaluation; HTN: hypertension; IC: intermittent claudication; RCT: randomized controlled trial; T2DM: type 2 diabetes mellitus. Open in new tab Characteristics of RCTs 1. Coronary heart disease Eight eligible RCTs (n = 1478 patients) conducted in Belgium,18,20 Canada,16,17 Denmark,19,23 Finland21 and The Netherlands,22 with follow-up ranging from six to 24 months were published between 2008 and 2018 (Table 2). Supervised exercise-based cardiac rehabilitation (SCR), distributed SCR, telemedicine cardiac rehabilitation (TCR), home-based cardiac rehabilitation (HBCR) or shared care cardiac rehabilitation (a general practitioner at a municipal clinic) interventions were compared with either usual care16,21 or SCR.17–20,22,23 Table 2. Exercise therapy, exercise training and exercise-based cardiac rehabilitation randomized controlled trials eligible for inclusion in the systematic review following full text screening. Reference . Country . Intervention . Comparator . Eco evala follow-up [months] . Coronary heart disease Oldridge et al.16 Canada Supervised CR (n = 93; 12%♀) Usual care (n = 95; 11%♀) 12 Papadakis et al.17 Canada Supervised distributed CR (n = 196; 16%♀) Supervised CR (n = 196; 11%♀) 24 Frederix et al.18 Belgium Telemedicine CR (n = 62; 16%♀) Supervised CR (n = 64; 20%♀) 6 Kidholm et al.19 Denmark Telemedicine CR (n = 72; 22%♀) Supervised CR (n = 69; 20%♀) 12 Frederix et al.20 Belgium Telemedicine CR follow-up (n = 62; 16%♀) Supervised CR (n = 64; 20%♀) 24 Hautala et al.21 Finland Supervised CR (n = 109; 27%♀) Usual care (n = 95; 29%♀) 12 Kraal et al.22 The Netherlands Home CR (n = 45; 11%♀) Supervised CR (n = 45; 11%♀) 12 Bertelsen et al.23 Denmark Shared care** CR (n = 106; 29%♀) Supervised CR (n = 106; 21%♀) 12 Chronic heart failure Reed et al.24 International (28%♀) Supervised exercise training (n = 1172) Usual care (n = 1159) 30 Intermittent claudication Spronk et al.25 The Netherlands Endovascular revascularization (n = 75; 59%♀) Supervised exercise training (n = 75; 53%♀) 12 Van Asselt et al.26 The Netherlands Supervised exercise training (n = 93; 45%♀) Walking advice (n = 83; 27%♀) 12 Mazari et al.27 UK Supervised exercise training (n = 60; 38%♀) Percutaneous transluminal angioplasty (n = 60; 38%♀) Supervised exercise training + Percutaneous transluminal angioplasty (n = 58; 33%♀) 12 Reynolds et al.28 USA Supervised exercise training (n = 37; 51%♀) Stent (n = 41; 30%♀) Optimum medical care (n = 20; 27%♀) 18 Body mass index ≥25 kg/m2 Gusi et al.29 Spain Supervised walking (n = 51; 100%♀) Usual care (n = 55; 100%♀) 6 Barton et al.30 UK (66%♀) Quadriceps strength exercise (n = 82) Quadriceps strength exercise + diet (n = 109) Usual care (leaflet with no exercise or diet advice) (n = 76) 24 Hypertension No RCT with ICER, ICUR or probability of cost-effectiveness reported Type 2 diabetes mellitus No RCT with ICER, ICUR or probability of cost-effectiveness reported Reference . Country . Intervention . Comparator . Eco evala follow-up [months] . Coronary heart disease Oldridge et al.16 Canada Supervised CR (n = 93; 12%♀) Usual care (n = 95; 11%♀) 12 Papadakis et al.17 Canada Supervised distributed CR (n = 196; 16%♀) Supervised CR (n = 196; 11%♀) 24 Frederix et al.18 Belgium Telemedicine CR (n = 62; 16%♀) Supervised CR (n = 64; 20%♀) 6 Kidholm et al.19 Denmark Telemedicine CR (n = 72; 22%♀) Supervised CR (n = 69; 20%♀) 12 Frederix et al.20 Belgium Telemedicine CR follow-up (n = 62; 16%♀) Supervised CR (n = 64; 20%♀) 24 Hautala et al.21 Finland Supervised CR (n = 109; 27%♀) Usual care (n = 95; 29%♀) 12 Kraal et al.22 The Netherlands Home CR (n = 45; 11%♀) Supervised CR (n = 45; 11%♀) 12 Bertelsen et al.23 Denmark Shared care** CR (n = 106; 29%♀) Supervised CR (n = 106; 21%♀) 12 Chronic heart failure Reed et al.24 International (28%♀) Supervised exercise training (n = 1172) Usual care (n = 1159) 30 Intermittent claudication Spronk et al.25 The Netherlands Endovascular revascularization (n = 75; 59%♀) Supervised exercise training (n = 75; 53%♀) 12 Van Asselt et al.26 The Netherlands Supervised exercise training (n = 93; 45%♀) Walking advice (n = 83; 27%♀) 12 Mazari et al.27 UK Supervised exercise training (n = 60; 38%♀) Percutaneous transluminal angioplasty (n = 60; 38%♀) Supervised exercise training + Percutaneous transluminal angioplasty (n = 58; 33%♀) 12 Reynolds et al.28 USA Supervised exercise training (n = 37; 51%♀) Stent (n = 41; 30%♀) Optimum medical care (n = 20; 27%♀) 18 Body mass index ≥25 kg/m2 Gusi et al.29 Spain Supervised walking (n = 51; 100%♀) Usual care (n = 55; 100%♀) 6 Barton et al.30 UK (66%♀) Quadriceps strength exercise (n = 82) Quadriceps strength exercise + diet (n = 109) Usual care (leaflet with no exercise or diet advice) (n = 76) 24 Hypertension No RCT with ICER, ICUR or probability of cost-effectiveness reported Type 2 diabetes mellitus No RCT with ICER, ICUR or probability of cost-effectiveness reported CR: cardiac rehabilitation; ICER: incremental cost-effectiveness ratio; ICUR: incremental cost–utility ratio; RCT: randomized controlled trial. aEco eval, economic evaluation. bShared care, municipal clinic + GP. Open in new tab Table 2. Exercise therapy, exercise training and exercise-based cardiac rehabilitation randomized controlled trials eligible for inclusion in the systematic review following full text screening. Reference . Country . Intervention . Comparator . Eco evala follow-up [months] . Coronary heart disease Oldridge et al.16 Canada Supervised CR (n = 93; 12%♀) Usual care (n = 95; 11%♀) 12 Papadakis et al.17 Canada Supervised distributed CR (n = 196; 16%♀) Supervised CR (n = 196; 11%♀) 24 Frederix et al.18 Belgium Telemedicine CR (n = 62; 16%♀) Supervised CR (n = 64; 20%♀) 6 Kidholm et al.19 Denmark Telemedicine CR (n = 72; 22%♀) Supervised CR (n = 69; 20%♀) 12 Frederix et al.20 Belgium Telemedicine CR follow-up (n = 62; 16%♀) Supervised CR (n = 64; 20%♀) 24 Hautala et al.21 Finland Supervised CR (n = 109; 27%♀) Usual care (n = 95; 29%♀) 12 Kraal et al.22 The Netherlands Home CR (n = 45; 11%♀) Supervised CR (n = 45; 11%♀) 12 Bertelsen et al.23 Denmark Shared care** CR (n = 106; 29%♀) Supervised CR (n = 106; 21%♀) 12 Chronic heart failure Reed et al.24 International (28%♀) Supervised exercise training (n = 1172) Usual care (n = 1159) 30 Intermittent claudication Spronk et al.25 The Netherlands Endovascular revascularization (n = 75; 59%♀) Supervised exercise training (n = 75; 53%♀) 12 Van Asselt et al.26 The Netherlands Supervised exercise training (n = 93; 45%♀) Walking advice (n = 83; 27%♀) 12 Mazari et al.27 UK Supervised exercise training (n = 60; 38%♀) Percutaneous transluminal angioplasty (n = 60; 38%♀) Supervised exercise training + Percutaneous transluminal angioplasty (n = 58; 33%♀) 12 Reynolds et al.28 USA Supervised exercise training (n = 37; 51%♀) Stent (n = 41; 30%♀) Optimum medical care (n = 20; 27%♀) 18 Body mass index ≥25 kg/m2 Gusi et al.29 Spain Supervised walking (n = 51; 100%♀) Usual care (n = 55; 100%♀) 6 Barton et al.30 UK (66%♀) Quadriceps strength exercise (n = 82) Quadriceps strength exercise + diet (n = 109) Usual care (leaflet with no exercise or diet advice) (n = 76) 24 Hypertension No RCT with ICER, ICUR or probability of cost-effectiveness reported Type 2 diabetes mellitus No RCT with ICER, ICUR or probability of cost-effectiveness reported Reference . Country . Intervention . Comparator . Eco evala follow-up [months] . Coronary heart disease Oldridge et al.16 Canada Supervised CR (n = 93; 12%♀) Usual care (n = 95; 11%♀) 12 Papadakis et al.17 Canada Supervised distributed CR (n = 196; 16%♀) Supervised CR (n = 196; 11%♀) 24 Frederix et al.18 Belgium Telemedicine CR (n = 62; 16%♀) Supervised CR (n = 64; 20%♀) 6 Kidholm et al.19 Denmark Telemedicine CR (n = 72; 22%♀) Supervised CR (n = 69; 20%♀) 12 Frederix et al.20 Belgium Telemedicine CR follow-up (n = 62; 16%♀) Supervised CR (n = 64; 20%♀) 24 Hautala et al.21 Finland Supervised CR (n = 109; 27%♀) Usual care (n = 95; 29%♀) 12 Kraal et al.22 The Netherlands Home CR (n = 45; 11%♀) Supervised CR (n = 45; 11%♀) 12 Bertelsen et al.23 Denmark Shared care** CR (n = 106; 29%♀) Supervised CR (n = 106; 21%♀) 12 Chronic heart failure Reed et al.24 International (28%♀) Supervised exercise training (n = 1172) Usual care (n = 1159) 30 Intermittent claudication Spronk et al.25 The Netherlands Endovascular revascularization (n = 75; 59%♀) Supervised exercise training (n = 75; 53%♀) 12 Van Asselt et al.26 The Netherlands Supervised exercise training (n = 93; 45%♀) Walking advice (n = 83; 27%♀) 12 Mazari et al.27 UK Supervised exercise training (n = 60; 38%♀) Percutaneous transluminal angioplasty (n = 60; 38%♀) Supervised exercise training + Percutaneous transluminal angioplasty (n = 58; 33%♀) 12 Reynolds et al.28 USA Supervised exercise training (n = 37; 51%♀) Stent (n = 41; 30%♀) Optimum medical care (n = 20; 27%♀) 18 Body mass index ≥25 kg/m2 Gusi et al.29 Spain Supervised walking (n = 51; 100%♀) Usual care (n = 55; 100%♀) 6 Barton et al.30 UK (66%♀) Quadriceps strength exercise (n = 82) Quadriceps strength exercise + diet (n = 109) Usual care (leaflet with no exercise or diet advice) (n = 76) 24 Hypertension No RCT with ICER, ICUR or probability of cost-effectiveness reported Type 2 diabetes mellitus No RCT with ICER, ICUR or probability of cost-effectiveness reported CR: cardiac rehabilitation; ICER: incremental cost-effectiveness ratio; ICUR: incremental cost–utility ratio; RCT: randomized controlled trial. aEco eval, economic evaluation. bShared care, municipal clinic + GP. Open in new tab 2. Chronic heart failure One eligible RCT (n = 2331 patients; stable, New York Heart Association class II–IV with left ventricular ejection fraction ≤35%) was conducted internationally (USA, 89%; Canada, 8%; France, 3%) with a follow-up of 30 months and published in 2010 with supervised exercise training (SET) compared with usual care (Table 2).24 3. Intermittent claudication Four eligible RCTs with SET (n = 602 patients) conducted in The Netherlands,25,26 the UK27 and the USA,28 with follow-up ranging from 12 to 18 months were published between 2008 and 2014 (Table 2). SET was compared with either endovascular revascularization,25 percutaneous transluminal angioplasty (PTA) alone or PTA plus SET,27 walking advice26 and either stenting or optimum medical care (OMC).28 4. Body mass index ≥25 kg/m2 Two RCTs (n = 373 patients) conducted in Spain29 and the UK30 with follow-up for six or 24 months were published between 2008 and 2009 (Table 2). Supervised walking exercise29 and quadriceps strength exercise (QSE) and QSE plus dietary intervention (QSE + diet) were compared with usual care.30 5. Hypertension No eligible title was identified. 6. Type 2 diabetes mellitus No eligible title was identified. Economic evaluation A. Perspective The healthcare perspective was used in four RCTs in patients CHD,16,17,19,21 one in patients with intermittent claudication27 and two in patients with BMI ≥25 kg/m2.29,30 The societal perspective was used in four RCTs in patients CHD,18,20,22,23 one in patients with CHF24 and three in patients with intermittent claudication (Table 3).25,26,28 Table 3. Perspective, incremental costs, incremental health benefits, incremental cost–utility ratios as of date given in original paper, and probability of cost-effectiveness with exercise therapy, exercise training and exercise-based cardiac rehabilitation randomized controlled trials in patients with coronary heart disease, chronic heart failure, intermittent claudication and body mass index ≥25 kg/m2. The italicized incremental cost–utility ratios were converted to US$ using the exchange rate for the year provided in each paper. Reference . Intervention/ comparator . Perspective . Mean incremental costs (currency date) . Mean incremental health benefit (QALY measure) . ICUR . Probability of cost-effectiveness . Coronary heart disease Oldridge et al.16 SCR/UC Healthcare US$663 (2006) 0.011 (QWB) 0.040 (TTO) US$60,270/QALY US$16,580/QALY 58% and 83% @ US$100,000 Papadakis et al.17 Standard three-month SCR/distributed 12-month SCR Healthcare −US$103 (2004) 0.009 (TTO) −US$11,400/QALY Dominant 67% to 63% @ US$ 0 to 100,000/QALY Frederix et al.18 SCR + 24 week TCR/SCR Societal −€564 (2015) 0.026 (EQ-5D) −€21,707/QALY Dominant −US$26,300/QALY Not reported Kidholm et al.19 TCR/SCR Healthcare €1664 (2014) 0.004 (SF-6D) €518,280/QALY US$710,000/QALY Not reported Frederix et al.20 SCR + TCR/SCR Societal −€878 (2015) 0.22 (EQ-5D) −€3993/QALY Dominant −US$4800/QALY Not reported Hautala et al.21 SCR/UC Healthcare −€1083 (2015) 0.045 (15D) −€24,511/QALY Dominant −US$29,700/QALY 100% @ ‘ any value of willingness-to-pay’ Kraal et al.22 HBCR/SCR Societal −€3160 (2015) 0.01 (SF-6D) Not reported, but Calculated as −€316,000/QALY −US$360,000/QALY 97% - 75% @ €0 to 100,00/QALY Bertelsen et al.23 Shared care CR/SCR Societal €336 (2015) 0.02 (EQ-5D) Not reported, but calculated as €16,800/QALY US$19,650/QALY 59% @ €40,000/QALY Chronic heart failure Reed et al.24 SET/UC Societal US$1161 (2008) 0.03 (EQ-5D) Not reported, but calculated as US$38,700 73% @ US$50,000/QALY Intermittent claudication Spronk et al.25 SET/endorevasc Societal −€2318 (2005) −0.01 (EQ-5D) €231,800/QALY US$290,000/QALY Not reported Van Asselt et al.26 SET/walk advice Societal €1104 (2008) 0.038 (EQ-5D) €28,693/QALY US$42,200/QALY 64% @ €40,000/QALY Mazari et al.27 SET/PTA SET/PTA + SET Healthcare −£3435 (2010) −£3045 0.009 (EQ-5D) −0.02 (EQ-5D) −£381,690/QALY −US$614,500/QALY Dominant £152,260/QALY US$245,100/QALY Not reported Reynolds et al.28 SET/stent SET/OMC Societal −US$4786 (2011) US$4625 −0.04 (EQ-5D) 0.12 (EQ-5D) US$119,650/QALY US$38,540/QALY 60% @ ∼US$30,000 to 80,000 Body mass index ≥25 kg/m2 Gusi et al.29 Supervised walking/UC Healthcare €41 (2005) 0.132 (EQ-5D) €311/QALY US$400/QALY 90% @ €350/QALY Barton et al.30 QSE + diet/UC QSE/UC Healthcare £647 (2007) £246 0.062 (EQ-5D) 0.005 (EQ-5D) £10,400/QALY US$21,000/QALY £49,200/QALY US$97,900/QALY 23% @ UK£20,000/QALY Reference . Intervention/ comparator . Perspective . Mean incremental costs (currency date) . Mean incremental health benefit (QALY measure) . ICUR . Probability of cost-effectiveness . Coronary heart disease Oldridge et al.16 SCR/UC Healthcare US$663 (2006) 0.011 (QWB) 0.040 (TTO) US$60,270/QALY US$16,580/QALY 58% and 83% @ US$100,000 Papadakis et al.17 Standard three-month SCR/distributed 12-month SCR Healthcare −US$103 (2004) 0.009 (TTO) −US$11,400/QALY Dominant 67% to 63% @ US$ 0 to 100,000/QALY Frederix et al.18 SCR + 24 week TCR/SCR Societal −€564 (2015) 0.026 (EQ-5D) −€21,707/QALY Dominant −US$26,300/QALY Not reported Kidholm et al.19 TCR/SCR Healthcare €1664 (2014) 0.004 (SF-6D) €518,280/QALY US$710,000/QALY Not reported Frederix et al.20 SCR + TCR/SCR Societal −€878 (2015) 0.22 (EQ-5D) −€3993/QALY Dominant −US$4800/QALY Not reported Hautala et al.21 SCR/UC Healthcare −€1083 (2015) 0.045 (15D) −€24,511/QALY Dominant −US$29,700/QALY 100% @ ‘ any value of willingness-to-pay’ Kraal et al.22 HBCR/SCR Societal −€3160 (2015) 0.01 (SF-6D) Not reported, but Calculated as −€316,000/QALY −US$360,000/QALY 97% - 75% @ €0 to 100,00/QALY Bertelsen et al.23 Shared care CR/SCR Societal €336 (2015) 0.02 (EQ-5D) Not reported, but calculated as €16,800/QALY US$19,650/QALY 59% @ €40,000/QALY Chronic heart failure Reed et al.24 SET/UC Societal US$1161 (2008) 0.03 (EQ-5D) Not reported, but calculated as US$38,700 73% @ US$50,000/QALY Intermittent claudication Spronk et al.25 SET/endorevasc Societal −€2318 (2005) −0.01 (EQ-5D) €231,800/QALY US$290,000/QALY Not reported Van Asselt et al.26 SET/walk advice Societal €1104 (2008) 0.038 (EQ-5D) €28,693/QALY US$42,200/QALY 64% @ €40,000/QALY Mazari et al.27 SET/PTA SET/PTA + SET Healthcare −£3435 (2010) −£3045 0.009 (EQ-5D) −0.02 (EQ-5D) −£381,690/QALY −US$614,500/QALY Dominant £152,260/QALY US$245,100/QALY Not reported Reynolds et al.28 SET/stent SET/OMC Societal −US$4786 (2011) US$4625 −0.04 (EQ-5D) 0.12 (EQ-5D) US$119,650/QALY US$38,540/QALY 60% @ ∼US$30,000 to 80,000 Body mass index ≥25 kg/m2 Gusi et al.29 Supervised walking/UC Healthcare €41 (2005) 0.132 (EQ-5D) €311/QALY US$400/QALY 90% @ €350/QALY Barton et al.30 QSE + diet/UC QSE/UC Healthcare £647 (2007) £246 0.062 (EQ-5D) 0.005 (EQ-5D) £10,400/QALY US$21,000/QALY £49,200/QALY US$97,900/QALY 23% @ UK£20,000/QALY CR: cardiac rehabilitation; endo revasc: endovascular revascularization; EQ-5D: EuroQoL-5D; HBCR: home-based cardiac rehabilitation; ICUR: incremental cost–utility ratio; OMC: optimum medical care; PTA: percutaneous transluminal angioplasty; QALY: quality-adjusted life-years; QSE: quadriceps strength exercise; QWB: Quality of Well-Being; SCR: supervised cardiac rehabilitation; SET: supervised exercise training; SF-6D: Short-Form-6 Dimension; TCR: telemedicine cardiac rehabilitation; TTO: time trade-off; UC: usual care; 15D, health-related quality of life. Open in new tab Table 3. Perspective, incremental costs, incremental health benefits, incremental cost–utility ratios as of date given in original paper, and probability of cost-effectiveness with exercise therapy, exercise training and exercise-based cardiac rehabilitation randomized controlled trials in patients with coronary heart disease, chronic heart failure, intermittent claudication and body mass index ≥25 kg/m2. The italicized incremental cost–utility ratios were converted to US$ using the exchange rate for the year provided in each paper. Reference . Intervention/ comparator . Perspective . Mean incremental costs (currency date) . Mean incremental health benefit (QALY measure) . ICUR . Probability of cost-effectiveness . Coronary heart disease Oldridge et al.16 SCR/UC Healthcare US$663 (2006) 0.011 (QWB) 0.040 (TTO) US$60,270/QALY US$16,580/QALY 58% and 83% @ US$100,000 Papadakis et al.17 Standard three-month SCR/distributed 12-month SCR Healthcare −US$103 (2004) 0.009 (TTO) −US$11,400/QALY Dominant 67% to 63% @ US$ 0 to 100,000/QALY Frederix et al.18 SCR + 24 week TCR/SCR Societal −€564 (2015) 0.026 (EQ-5D) −€21,707/QALY Dominant −US$26,300/QALY Not reported Kidholm et al.19 TCR/SCR Healthcare €1664 (2014) 0.004 (SF-6D) €518,280/QALY US$710,000/QALY Not reported Frederix et al.20 SCR + TCR/SCR Societal −€878 (2015) 0.22 (EQ-5D) −€3993/QALY Dominant −US$4800/QALY Not reported Hautala et al.21 SCR/UC Healthcare −€1083 (2015) 0.045 (15D) −€24,511/QALY Dominant −US$29,700/QALY 100% @ ‘ any value of willingness-to-pay’ Kraal et al.22 HBCR/SCR Societal −€3160 (2015) 0.01 (SF-6D) Not reported, but Calculated as −€316,000/QALY −US$360,000/QALY 97% - 75% @ €0 to 100,00/QALY Bertelsen et al.23 Shared care CR/SCR Societal €336 (2015) 0.02 (EQ-5D) Not reported, but calculated as €16,800/QALY US$19,650/QALY 59% @ €40,000/QALY Chronic heart failure Reed et al.24 SET/UC Societal US$1161 (2008) 0.03 (EQ-5D) Not reported, but calculated as US$38,700 73% @ US$50,000/QALY Intermittent claudication Spronk et al.25 SET/endorevasc Societal −€2318 (2005) −0.01 (EQ-5D) €231,800/QALY US$290,000/QALY Not reported Van Asselt et al.26 SET/walk advice Societal €1104 (2008) 0.038 (EQ-5D) €28,693/QALY US$42,200/QALY 64% @ €40,000/QALY Mazari et al.27 SET/PTA SET/PTA + SET Healthcare −£3435 (2010) −£3045 0.009 (EQ-5D) −0.02 (EQ-5D) −£381,690/QALY −US$614,500/QALY Dominant £152,260/QALY US$245,100/QALY Not reported Reynolds et al.28 SET/stent SET/OMC Societal −US$4786 (2011) US$4625 −0.04 (EQ-5D) 0.12 (EQ-5D) US$119,650/QALY US$38,540/QALY 60% @ ∼US$30,000 to 80,000 Body mass index ≥25 kg/m2 Gusi et al.29 Supervised walking/UC Healthcare €41 (2005) 0.132 (EQ-5D) €311/QALY US$400/QALY 90% @ €350/QALY Barton et al.30 QSE + diet/UC QSE/UC Healthcare £647 (2007) £246 0.062 (EQ-5D) 0.005 (EQ-5D) £10,400/QALY US$21,000/QALY £49,200/QALY US$97,900/QALY 23% @ UK£20,000/QALY Reference . Intervention/ comparator . Perspective . Mean incremental costs (currency date) . Mean incremental health benefit (QALY measure) . ICUR . Probability of cost-effectiveness . Coronary heart disease Oldridge et al.16 SCR/UC Healthcare US$663 (2006) 0.011 (QWB) 0.040 (TTO) US$60,270/QALY US$16,580/QALY 58% and 83% @ US$100,000 Papadakis et al.17 Standard three-month SCR/distributed 12-month SCR Healthcare −US$103 (2004) 0.009 (TTO) −US$11,400/QALY Dominant 67% to 63% @ US$ 0 to 100,000/QALY Frederix et al.18 SCR + 24 week TCR/SCR Societal −€564 (2015) 0.026 (EQ-5D) −€21,707/QALY Dominant −US$26,300/QALY Not reported Kidholm et al.19 TCR/SCR Healthcare €1664 (2014) 0.004 (SF-6D) €518,280/QALY US$710,000/QALY Not reported Frederix et al.20 SCR + TCR/SCR Societal −€878 (2015) 0.22 (EQ-5D) −€3993/QALY Dominant −US$4800/QALY Not reported Hautala et al.21 SCR/UC Healthcare −€1083 (2015) 0.045 (15D) −€24,511/QALY Dominant −US$29,700/QALY 100% @ ‘ any value of willingness-to-pay’ Kraal et al.22 HBCR/SCR Societal −€3160 (2015) 0.01 (SF-6D) Not reported, but Calculated as −€316,000/QALY −US$360,000/QALY 97% - 75% @ €0 to 100,00/QALY Bertelsen et al.23 Shared care CR/SCR Societal €336 (2015) 0.02 (EQ-5D) Not reported, but calculated as €16,800/QALY US$19,650/QALY 59% @ €40,000/QALY Chronic heart failure Reed et al.24 SET/UC Societal US$1161 (2008) 0.03 (EQ-5D) Not reported, but calculated as US$38,700 73% @ US$50,000/QALY Intermittent claudication Spronk et al.25 SET/endorevasc Societal −€2318 (2005) −0.01 (EQ-5D) €231,800/QALY US$290,000/QALY Not reported Van Asselt et al.26 SET/walk advice Societal €1104 (2008) 0.038 (EQ-5D) €28,693/QALY US$42,200/QALY 64% @ €40,000/QALY Mazari et al.27 SET/PTA SET/PTA + SET Healthcare −£3435 (2010) −£3045 0.009 (EQ-5D) −0.02 (EQ-5D) −£381,690/QALY −US$614,500/QALY Dominant £152,260/QALY US$245,100/QALY Not reported Reynolds et al.28 SET/stent SET/OMC Societal −US$4786 (2011) US$4625 −0.04 (EQ-5D) 0.12 (EQ-5D) US$119,650/QALY US$38,540/QALY 60% @ ∼US$30,000 to 80,000 Body mass index ≥25 kg/m2 Gusi et al.29 Supervised walking/UC Healthcare €41 (2005) 0.132 (EQ-5D) €311/QALY US$400/QALY 90% @ €350/QALY Barton et al.30 QSE + diet/UC QSE/UC Healthcare £647 (2007) £246 0.062 (EQ-5D) 0.005 (EQ-5D) £10,400/QALY US$21,000/QALY £49,200/QALY US$97,900/QALY 23% @ UK£20,000/QALY CR: cardiac rehabilitation; endo revasc: endovascular revascularization; EQ-5D: EuroQoL-5D; HBCR: home-based cardiac rehabilitation; ICUR: incremental cost–utility ratio; OMC: optimum medical care; PTA: percutaneous transluminal angioplasty; QALY: quality-adjusted life-years; QSE: quadriceps strength exercise; QWB: Quality of Well-Being; SCR: supervised cardiac rehabilitation; SET: supervised exercise training; SF-6D: Short-Form-6 Dimension; TCR: telemedicine cardiac rehabilitation; TTO: time trade-off; UC: usual care; 15D, health-related quality of life. Open in new tab B. Costs 1. Coronary heart disease Costs were both higher16 and lower21 with SCR compared with usual care (Table 3). Higher costs were reported with TCR19 and shared care cardiac rehabilitation23 than with SCR. Lower costs were reported with a standard three-month SCR compared with a 12-month distributed SCR17 and with SCR + TCR18,20 and HBCR22 compared with SCR. 2. Chronic heart failure Costs were higher with SET compared with usual care (Table 3).24 3. Intermittent claudication Higher costs were reported with SET than with walking advice26 OMC (Table 3).28 Lower costs were reported with SET compared with endovascular revascularization,25 either PTA or PTA + SET27 and stenting (Table 3).28 4. Body mass index ≥25 kg/m2 Higher costs were reported with supervised walking29 and both QSE and QSE + diet30 than with usual care (Table 3). C. Health benefits Walking distance was reported in one RCT26 and generic health measures were used in the other papers to calculate the QALYs gained (Table 3). 1. Coronary heart disease Regardless of the measure used to estimate QALYs, more QALYs were gained with SCR than with either usual care,16,21 shared care,23 TCR,18–20 HBCR22 and distributed SCR17 than with SCR. 2. Chronic heart failure More QALYs were gained with SET than with usual care.24 3. Intermittent claudication More QALYs were gained with SET than with either walking advice26 or OMC.28 Fewer QALYs were gained with SET than with either endovascular revascularization25 or stenting.28 There was no statistical difference in the QALYs gained with PTA, SET or PTA + SET.27 4. Body mass index ≥25 kg/m2 More QALYs were gained with supervised walking29 and with QSE or QSE + diet30 than with usual care. D. Incremental cost-effectiveness ratios 1. Intermittent claudication A single ICER, €4.08/m, was reported with SET compared with walking advice.26 E. Incremental cost–utility ratios and probability of cost-effectiveness When an ICUR was not reported, the reason given was typically that the difference between intervention costs and/or health benefits was not significant (Table 3). 1. Coronary heart disease The seven ICURs ranged from €518,280/QALY to −€3993/QALY. Four ICURs were reported to be dominant, one with SCR compared with distributed SCR,17 two with SCR + TCR18,20 and one with TCR21 compared with SCR. The probability of cost-effectiveness ranged from 58%16 to 100%21 with SCR compared with usual care. 2. Chronic heart failure An ICUR of US$38,700 with SET compared with usual care was calculated from the reported data in the RCT with a reported 73% probability of cost-effectiveness.24 3. Intermittent claudication The ICURs with SET ranged from €28,693/QALY26 with walking advice to £231,800/QALY with endovascular revescularization.25 The ICUR with SET compared with PTA was −£381,694/QALY (dominant) and £152,260/QALY compared with PTA + SET.27 The ICURs with SET ranged from US$38,541/QALY compared with OMC to US$119,650/QALY compared with stenting.28 The probability of SET being cost-effective compared with endovascular revascularization was not reported,25 60% with SET compared with stenting28 and 64% with SET compared with walk advice.26 4. Body mass index ≥ 25 kg/m2 Supervised walking was cost-effective when compared with usual care in overweight or obese and depressed women.29 Compared with usual care, QSE + diet was cost-effective with QSE alone subject to ‘extended dominance’ as both QSE + diet and usual care ‘could provide higher benefit at equivalent cost’.30 Compared with usual care, the probability of cost-effectiveness ranged from 23% with QSE + diet30 to 90% with supervised walking.29 Value for money Both the WHO guideline10 and the ACC/AHA guideline,12 the latter with all Euro and UK £ICURs expressed as US$, were used to determine the value for money of prescribed exercise therapy (Table 4). Table 4. Incremental cost–utility ratios in euros or US$ with World Health Organization and American College of Cardiology/American Heart Association value for money guidelines with exercise therapy, exercise training and exercise-based cardiac rehabilitation randomized controlled trials in patients with coronary heart disease, chronic heart failure, intermittent claudication and body mass index ≥25. The italicized incremental cost–utility ratios were converted to US$ using the exchange rate for the year provided in each paper for the American College of Cardiology/American Heart Association guidelines. Reference . Intervention/comparator . ICUR/QALY . WHOa guideline . ACC/AHAb guideline . Coronary heart disease Oldridge et al.16 SCR/UC US$60,270/QWB QALY US$16,580/TTO QALY CE Highly CE Intermediate value High value Papadakis et al.17 3-month SCR/ 12-month DSCR US$-11,400/TTO QALY Highly CE High value Frederix et al.18 SCR + 24-week TCR/SCR −€21,707/EQ-5DQALY −US$26,300/EQ-5DQALY Highly CE High value Kidholm et al.19 TCR/SCR €518,280/SF-6 QALY US$710,000/SF-6 QALY Not CE Low value Frederix et al.20 SCR + TCR/SCRc −€3993/EQ-5D QALY −US$4800/EQ-5D QALY Highly CE High value Hautala et al.21 SCR/UC −€24,511/15D QALY −US$29,700/15D QALY Highly CE High value Kraal et al.22 HBCR/SCR −€316,000/SF-6D QALYd −US$360,000/SF-6D QALY Highly CE High value Bertelsen et al.23 Shared care CR/SCR €16,800/EQ-5D QALYd US$19,650/EQ-5D QALY Highly CE High value Chronic heart failure Reed et al.24 SET/UC US$38,700/EQ-5D QALYd Highly CE High value Intermittent claudication Spronk et al.25 SET/endo revasc €231,800/EQ-5D QALY US$290,000/EQ-5D QALY Not CE Low value Van Asselt et al.26 SET/walk advice €28,693/EQ-5D QALY US$42,200/EQ-5D QALY Highly CE High value Mazari et al.27 SET/PTA SET/PTA SET −£381,690/EQ-5D QALY −US$614,500/EQ-5D QALY £152,260/EQ-5D QALY US$245,100/EQ-5D QALY Not CE Not CE Low value Low value Reynolds et al.28,e SET/stent SET/OMC US$119,650/EQ-5D QALY US$38,540/EQ-5D QALY Not CE Highly CE Low value High value Body mass index ≥25 kg/m2 Gusi et al.29 Supervised walking/UC €311/EQ-5D QALY US$400/EQ-5D QALY Highly CE High value Barton et al.30 QSE + diet/UC QSE/UC £10,400/EQ-5D QALY US$21,000/EQ-5D QALY £49,200/EQ-5D QALY US$97,900/EQ-5D QALY Highly CE CE High value Intermediate value Reference . Intervention/comparator . ICUR/QALY . WHOa guideline . ACC/AHAb guideline . Coronary heart disease Oldridge et al.16 SCR/UC US$60,270/QWB QALY US$16,580/TTO QALY CE Highly CE Intermediate value High value Papadakis et al.17 3-month SCR/ 12-month DSCR US$-11,400/TTO QALY Highly CE High value Frederix et al.18 SCR + 24-week TCR/SCR −€21,707/EQ-5DQALY −US$26,300/EQ-5DQALY Highly CE High value Kidholm et al.19 TCR/SCR €518,280/SF-6 QALY US$710,000/SF-6 QALY Not CE Low value Frederix et al.20 SCR + TCR/SCRc −€3993/EQ-5D QALY −US$4800/EQ-5D QALY Highly CE High value Hautala et al.21 SCR/UC −€24,511/15D QALY −US$29,700/15D QALY Highly CE High value Kraal et al.22 HBCR/SCR −€316,000/SF-6D QALYd −US$360,000/SF-6D QALY Highly CE High value Bertelsen et al.23 Shared care CR/SCR €16,800/EQ-5D QALYd US$19,650/EQ-5D QALY Highly CE High value Chronic heart failure Reed et al.24 SET/UC US$38,700/EQ-5D QALYd Highly CE High value Intermittent claudication Spronk et al.25 SET/endo revasc €231,800/EQ-5D QALY US$290,000/EQ-5D QALY Not CE Low value Van Asselt et al.26 SET/walk advice €28,693/EQ-5D QALY US$42,200/EQ-5D QALY Highly CE High value Mazari et al.27 SET/PTA SET/PTA SET −£381,690/EQ-5D QALY −US$614,500/EQ-5D QALY £152,260/EQ-5D QALY US$245,100/EQ-5D QALY Not CE Not CE Low value Low value Reynolds et al.28,e SET/stent SET/OMC US$119,650/EQ-5D QALY US$38,540/EQ-5D QALY Not CE Highly CE Low value High value Body mass index ≥25 kg/m2 Gusi et al.29 Supervised walking/UC €311/EQ-5D QALY US$400/EQ-5D QALY Highly CE High value Barton et al.30 QSE + diet/UC QSE/UC £10,400/EQ-5D QALY US$21,000/EQ-5D QALY £49,200/EQ-5D QALY US$97,900/EQ-5D QALY Highly CE CE High value Intermediate value ACC/AHA: American College of Cardiology/American Heart Association; CE: cost-effective; DSCR: distributed supervised cardiac rehabilitation; endo revasc: endovascular revascularization; EQ-5D: EuroQoL-5D; GDP: gross domestic product; HBCR: home-based cardiac rehabilitation; OMC: optimum medical care; PTA: percutaneous transluminal angioplasty; QSE: quadriceps strength exercise; QWB: Quality of Well-Being; QALY: quality-adjusted life-year; SCR: supervised cardiac rehabilitation; SET: supervised exercise training;SF-6D: Short-Form-6 Dimension; TCR: telemedicine cardiac rehabilitation; TTO: time trade-off; UC: usual care; WHO: World Health Organization; 15D, health-related quality of life. a WHO guideline (GDP per capita): highly CE, <GDP; CE, ≥GDP to three times GDP. b ACC/AHA guideline: high value, US$/QALY <US$50,000; intermediate value, US$50,000 to <US$150,000; low value, >US$150,000; not CE, more than three times GDP. c 24-month follow-up with no intervention. d Calculated ICUR. e Observed data. Open in new tab Table 4. Incremental cost–utility ratios in euros or US$ with World Health Organization and American College of Cardiology/American Heart Association value for money guidelines with exercise therapy, exercise training and exercise-based cardiac rehabilitation randomized controlled trials in patients with coronary heart disease, chronic heart failure, intermittent claudication and body mass index ≥25. The italicized incremental cost–utility ratios were converted to US$ using the exchange rate for the year provided in each paper for the American College of Cardiology/American Heart Association guidelines. Reference . Intervention/comparator . ICUR/QALY . WHOa guideline . ACC/AHAb guideline . Coronary heart disease Oldridge et al.16 SCR/UC US$60,270/QWB QALY US$16,580/TTO QALY CE Highly CE Intermediate value High value Papadakis et al.17 3-month SCR/ 12-month DSCR US$-11,400/TTO QALY Highly CE High value Frederix et al.18 SCR + 24-week TCR/SCR −€21,707/EQ-5DQALY −US$26,300/EQ-5DQALY Highly CE High value Kidholm et al.19 TCR/SCR €518,280/SF-6 QALY US$710,000/SF-6 QALY Not CE Low value Frederix et al.20 SCR + TCR/SCRc −€3993/EQ-5D QALY −US$4800/EQ-5D QALY Highly CE High value Hautala et al.21 SCR/UC −€24,511/15D QALY −US$29,700/15D QALY Highly CE High value Kraal et al.22 HBCR/SCR −€316,000/SF-6D QALYd −US$360,000/SF-6D QALY Highly CE High value Bertelsen et al.23 Shared care CR/SCR €16,800/EQ-5D QALYd US$19,650/EQ-5D QALY Highly CE High value Chronic heart failure Reed et al.24 SET/UC US$38,700/EQ-5D QALYd Highly CE High value Intermittent claudication Spronk et al.25 SET/endo revasc €231,800/EQ-5D QALY US$290,000/EQ-5D QALY Not CE Low value Van Asselt et al.26 SET/walk advice €28,693/EQ-5D QALY US$42,200/EQ-5D QALY Highly CE High value Mazari et al.27 SET/PTA SET/PTA SET −£381,690/EQ-5D QALY −US$614,500/EQ-5D QALY £152,260/EQ-5D QALY US$245,100/EQ-5D QALY Not CE Not CE Low value Low value Reynolds et al.28,e SET/stent SET/OMC US$119,650/EQ-5D QALY US$38,540/EQ-5D QALY Not CE Highly CE Low value High value Body mass index ≥25 kg/m2 Gusi et al.29 Supervised walking/UC €311/EQ-5D QALY US$400/EQ-5D QALY Highly CE High value Barton et al.30 QSE + diet/UC QSE/UC £10,400/EQ-5D QALY US$21,000/EQ-5D QALY £49,200/EQ-5D QALY US$97,900/EQ-5D QALY Highly CE CE High value Intermediate value Reference . Intervention/comparator . ICUR/QALY . WHOa guideline . ACC/AHAb guideline . Coronary heart disease Oldridge et al.16 SCR/UC US$60,270/QWB QALY US$16,580/TTO QALY CE Highly CE Intermediate value High value Papadakis et al.17 3-month SCR/ 12-month DSCR US$-11,400/TTO QALY Highly CE High value Frederix et al.18 SCR + 24-week TCR/SCR −€21,707/EQ-5DQALY −US$26,300/EQ-5DQALY Highly CE High value Kidholm et al.19 TCR/SCR €518,280/SF-6 QALY US$710,000/SF-6 QALY Not CE Low value Frederix et al.20 SCR + TCR/SCRc −€3993/EQ-5D QALY −US$4800/EQ-5D QALY Highly CE High value Hautala et al.21 SCR/UC −€24,511/15D QALY −US$29,700/15D QALY Highly CE High value Kraal et al.22 HBCR/SCR −€316,000/SF-6D QALYd −US$360,000/SF-6D QALY Highly CE High value Bertelsen et al.23 Shared care CR/SCR €16,800/EQ-5D QALYd US$19,650/EQ-5D QALY Highly CE High value Chronic heart failure Reed et al.24 SET/UC US$38,700/EQ-5D QALYd Highly CE High value Intermittent claudication Spronk et al.25 SET/endo revasc €231,800/EQ-5D QALY US$290,000/EQ-5D QALY Not CE Low value Van Asselt et al.26 SET/walk advice €28,693/EQ-5D QALY US$42,200/EQ-5D QALY Highly CE High value Mazari et al.27 SET/PTA SET/PTA SET −£381,690/EQ-5D QALY −US$614,500/EQ-5D QALY £152,260/EQ-5D QALY US$245,100/EQ-5D QALY Not CE Not CE Low value Low value Reynolds et al.28,e SET/stent SET/OMC US$119,650/EQ-5D QALY US$38,540/EQ-5D QALY Not CE Highly CE Low value High value Body mass index ≥25 kg/m2 Gusi et al.29 Supervised walking/UC €311/EQ-5D QALY US$400/EQ-5D QALY Highly CE High value Barton et al.30 QSE + diet/UC QSE/UC £10,400/EQ-5D QALY US$21,000/EQ-5D QALY £49,200/EQ-5D QALY US$97,900/EQ-5D QALY Highly CE CE High value Intermediate value ACC/AHA: American College of Cardiology/American Heart Association; CE: cost-effective; DSCR: distributed supervised cardiac rehabilitation; endo revasc: endovascular revascularization; EQ-5D: EuroQoL-5D; GDP: gross domestic product; HBCR: home-based cardiac rehabilitation; OMC: optimum medical care; PTA: percutaneous transluminal angioplasty; QSE: quadriceps strength exercise; QWB: Quality of Well-Being; QALY: quality-adjusted life-year; SCR: supervised cardiac rehabilitation; SET: supervised exercise training;SF-6D: Short-Form-6 Dimension; TCR: telemedicine cardiac rehabilitation; TTO: time trade-off; UC: usual care; WHO: World Health Organization; 15D, health-related quality of life. a WHO guideline (GDP per capita): highly CE, <GDP; CE, ≥GDP to three times GDP. b ACC/AHA guideline: high value, US$/QALY <US$50,000; intermediate value, US$50,000 to <US$150,000; low value, >US$150,000; not CE, more than three times GDP. c 24-month follow-up with no intervention. d Calculated ICUR. e Observed data. Open in new tab 1. Coronary heart disease Seven of the nine exercise-based cardiac rehabilitation ICURs16–18,20–23 were ‘highly cost-effective’ or ‘high value’ (two of which were calculated),22,23 one was ‘cost-effective’ or ‘intermediate value’16 and ‘not cost-effective’ one was ‘low value.’19 2. Chronic heart failure The calculated ICUR in the single RCT in patients with heart failure was consistent with SET being ‘highly cost-effective’ or ‘high value’ compared with usual care.24 3. Intermittent claudication The ICURs with SET compared with walking advice26 and SET compared with OMC28 were both rated as ‘highly cost-effective’ or ‘high value’. The ICURs with SET compared with endovascular revascularization,25 with SET compared with either PTA alone or PTA + SET27 and with SET compared with stenting28 were all rated as ‘not cost-effective’ or ‘low value’. 4. Body mass index ≥25 The ICURs with supervised walking29 and with QSE and QSE + diet30 were rated as ‘highly cost-effective’ or ‘high value’ compared with usual care. Discussion There is strong RCT evidence for the effectiveness of prescribed supervised exercise therapy, exercise training and exercise-based cardiac rehabilitation in patients with CHD, CHF, intermittent claudication and the cardiovascular risk factors of BMI ≥25 kg/m2, hypertension and T2DM.4,5 However, economic evaluations conducted alongside RCTs are limited.6–8 A systematic search of seven electronic databases identified 15 economic evaluations conducted alongside prescribed supervised exercise therapy, exercise training and exercise-based cardiac rehabilitation RCTs in patients with CHD, CHF, intermittent claudication or BMI ≥25 kg/m2,16–30 with no eligible economic evaluation in patients with either hypertension or T2DM. Of the 19 ICURs reported, 63% were ‘highly cost-effective’ using the WHO < 1-times national GDP threshold10 or ‘high value’ using the ACC/AHA < US$50,000 threshold,12 with 26% reported as dominant. Of the nine ICURs in patients with CHD, 78% were ‘highly cost-effective’/‘high value’ with 44% dominant; the single ICUR in patients with CHF was ‘highly cost-effective’/‘high value’; 33% of six ICURs in patients with intermittent claudication were ‘highly cost-effective’/‘high value’ with one dominant; and 66% of two ICURs in patients with BMI ≥25 kg/m2 were ‘highly cost-effective’/‘high value’. However, with the introduction of new healthcare technologies, rapidly escalating costs and evolving healthcare and economic environments, more evidence for the cost-effectiveness of prescribed supervised exercise therapy, exercise training and exercise-based cardiac rehabilitation based on RCTs, especially in patients with hypertension or T2DM, is needed. The exercise therapy data in this review essentially substantiate observations that there is a ‘large variation’ in the data and that ‘some exercise interventions can be cost-effective’, but that even the most convincing evidence is ‘partly contradictory’6 and that the ‘physical activity interventions differed in terms of type, volume, and duration of exercise performed, even within disease categories’.7 The data also essentially substantiate reported concerns of ‘uncertainty due to study design and data, definitions for standard care and subgroup analyses’ specifically with regards to cardiac rehabilitation.8 There are a number of limitations to the data reviewed in this systematic review. With the WHO guideline criticized as having major shortcomings11 and with research suggesting that the ACC/AHA ‘highly cost-effective’ threshold may be too low,31 the proportion of ‘highly cost-effective’ ICURs in this review may well be over-estimated. Between-diagnosis ICUR comparisons were not attempted as the intervention arms, the comparator arms and the instruments used to estimate QALYs varied widely across the RCTs. Further, the variation in the reported ICURs was large, ranging from US$710,000/SF-6D QALY20 to −US$360,00/SF-6D QALY22 in patients with CHD, from US$290,000/EQ-5D QALY25 to −US$614,500/EQ-5D QALY27 in patients with IC and from US$97,900/EQ-5D QALY30 to US$400/EQ-5D/QALY29 in patients with BMI ≥25 kg/m2. The fact that all of the eligible RCTs were reported in English raises the possibility of bias; with the grey literature not considered, there also is the risk of missing unpublished RCTs with negative or uncertain results. The diagnosis considered in individual RCTs was not always ‘pure’ because there were populations with ‘mixed’ diagnoses, which may also have introduced some bias. For example, the CHD populations recruited included patients with a diagnosis of myocardial infarction only,16 angina or myocardial infarction,17 CHD and/or CHF,18,20 and acute coronary syndrome.19,21–23 Confirmation with the CHEERS reporting standards varied widely across the included RCTs. The strengths of the data reported in this systematic review of prescribed exercise include the a priori specification that the economic evaluation data needed to have been collected alongside an RCT and that an ICER, an ICUR and/or the probability of cost-effectiveness needed to have been reported in each paper. Two common cost-effectiveness value for money thresholds10,12 were systematically applied to each paper. With prescribed exercise therapy, exercise training and exercise-based cardiac rehabilitation as the interventions, 63% of the 19 ICURs met the WHO and ACC/AHA cost-effectiveness threshold criteria as ‘highly cost-effective’ or ‘high value’ interventions. However, further cost-effectiveness research is needed, especially in patients with CHF, intermittent claudication or BMI ≥25 kg/m2, where the number of economic evaluations alongside an RCT is small, and in patients with either hypertension or T2DM where no economic evaluation alongside an RCT was identified. Author contribution NO contributed to the conception or design of the work. NO and RT contributed to the acquisition, analysis or interpretation of data for the work. NO drafted the manuscript. RT critically revised the manuscript. All gave final approval and agree to be accountable for all aspects of work ensuring integrity and accuracy. Acknowledgements The search strategies were developed in collaboration with an information specialist, Maureen Pakosh (Library & Information Services UHN, Toronto Rehab – Rumsey Cardiac Centre Library, Toronto, Canada), whose expertise is gratefully acknowledged. The initial systematic searches were conducted by Karam Turk-Adawi (University of Qatar, Doha, Qatar). The expertise and patience of the SAGE editorial staff is acknowledged. 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 Booth FW , Roberts CK, Thyfault JP, et al. Role of inactivity in chronic diseases: Evolutionary insight and pathophysiological mechanisms . Physiol Rev 2017 ; 97 : 1351 – 1402 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Ding D , Kolbe-Alexander T, Nguyen B, et al. The economic burden of physical inactivity: A systematic review and critical appraisal . Br J Sports Med 2017 ; 51 : 1392 – 1409 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Ding D , Lawson KD, Kolbe-Alexander TL, et al. The economic burden of physical inactivity: A global analysis of major non-communicable diseases . Lancet 2016 ; 388 : 1311 – 1324 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Pedersen BK , Saltin B. Evidence for prescribing exercise as therapy in chronic disease . Scand J Med Sci Sports 2006 ; 16 ( Suppl 1 ): 3 – 63 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Pedersen BK , Saltin B. Exercise as medicine – evidence for prescribing exercise as therapy in 26 different chronic diseases . Scand J Med Sci Sports 2015 ; 25 ( Suppl 3 ): 1 – 72 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Roine E , Roine RP, Rasanen P, et al. Cost-effectiveness of interventions based on physical exercise in the treatment of various diseases: A systematic literature review . Int J Technol Assess Health Care 2009 ; 25 : 427 – 454 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Guillon M , Rochaix L, Dupont JK. Cost-effectiveness of interventions based on physical activity in the treatment of chronic conditions: A systematic literature review . Int J Technol Assess Health Care 2018 ; 34 : 481 – 497 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Shields GE , Wells A, Doherty P, et al. Cost-effectiveness of cardiac rehabilitation: A systematic review . Heart 2018 ; 104 : 1403 – 1410 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Drummond MF, Sculpher MA, Claxton K, et al. Methods for the economic evaluation of health care programmes. 4th ed. Oxford: Oxford University Press, 2015 . 10 World Health Organization. The World Health Report 2002 – reducing risks, promoting healthy life, www.who.int/whr/2002/en/ (2002, accessed 20 June 2019) . 11 Marseille E , Larson B, Kazi DS, et al. Thresholds for the cost-effectiveness of interventions: Alternative approaches . Bull World Health Organ 2015 ; 93 : 118 – 124 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Anderson JL , Heidenreich PA, Barnett PG, et al. ACC/AHA statement on cost/value methodology in clinical practice guidelines and performance measures: A report of the American College of Cardiology/American Heart Association Task Force on Performance Measures and Task Force on Practice Guidelines . Circulation 2014 ; 129 : 2329 – 2345 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Moher D , Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement . BMJ 2009 ; 339 : b2535 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Higgins J and Green S (eds). Cochrane handbook for systematic reviews of interventions version 5.1.0 (updated March 2011). The Cochrane Collaboration, http://handbook-5-1.cochrane.org/) (2011, accessed 20 May, 2019) . 15 Husereau D , Drummond M, Petrou S, et al. Consolidated health economic evaluation reporting standards (CHEERS)—explanation and elaboration: A report of the ISPOR Health Economic Evaluation Publication Guidelines Good Reporting Practices Task Force . Value Health 2013 ; 16 : 231 – 250 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Oldridge N , Furlong W, Perkins A, et al. Community or patient preferences for cost-effectiveness of cardiac rehabilitation: Does it matter? . Eur J Cardiovasc Prev Rehabil 2008 ; 15 : 608 – 615 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Papadakis S , Reid RD, Coyle D, et al. Cost-effectiveness of cardiac rehabilitation program delivery models in patients at varying cardiac risk, reason for referral, and sex . Eur J Cardiovasc Prev Rehabil 2008 ; 15 : 347 – 353 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Frederix I , Hansen D, Coninx K, et al. Effect of comprehensive cardiac telerehabilitation on one-year cardiovascular rehospitalization rate, medical costs and quality of life: A cost-effectiveness analysis . Eur J Prev Cardiol 2016 ; 23 : 674 – 682 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Kidholm K , Rasmussen MK, Andreasen JJ, et al. Cost–utility analysis of a cardiac telerehabilitation program: The Teledialog Project . Telemed J E Health 2016 ; 22 : 553 – 563 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Frederix I , Solmi F, Piepoli MF, et al. Cardiac telerehabilitation: A novel cost-efficient care delivery strategy that can induce long-term health benefits . Eur J Prev Cardiol 2017 ; 24 : 1708 – 1717 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Hautala AJ , Kiviniemi AM, Makikallio T, et al. Economic evaluation of exercise-based cardiac rehabilitation in patients with a recent acute coronary syndrome . Scand J Med Sci Sports 2017 ; 27 : 1395 – 1403 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Kraal JJ , Elske Van Den Akker-Van Marle M, Abu-Hanna A, et al. Clinical and cost-effectiveness of home-based cardiac rehabilitation compared to conventional, centre-based cardiac rehabilitation: Results of the FIT@Home study . Eur J Prev Cardiol 2017 ; 24 : 1260 – 1273 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Bertelsen JB , Dehbarez NT, Refsgaard J, et al. Shared care versus hospital-based cardiac rehabilitation: A cost–utility analysis based on a randomised controlled trial . Open Heart 2018 ; 5 : e000584 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 24 Reed SD , Whellan DJ, Li Y, et al. Economic evaluation of the HF-ACTION (Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training) randomized controlled trial: An exercise training study of patients with chronic heart failure . Circ Cardiovasc Qual Outcomes 2010 ; 3 : 374 – 381 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Spronk S , Bosch JL, den Hoed PT, et al. Cost-effectiveness of endovascular revascularization compared to supervised hospital-based exercise training in patients with intermittent claudication: A randomized controlled trial . J Vasc Surg 2008 ; 48 : 1472 – 1480 . Google Scholar Crossref Search ADS PubMed WorldCat 26 van Asselt AD , Nicolai SP, Joore MA, et al. Cost-effectiveness of exercise therapy in patients with intermittent claudication: Supervised exercise therapy versus a ‘go home and walk’ advice . Eur J Vasc Endovasc Surg 2011 ; 41 : 97 – 103 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Mazari FA , Khan JA, Carradice D, et al. Economic analysis of a randomized trial of percutaneous angioplasty, supervised exercise or combined treatment for intermittent claudication due to femoropopliteal arterial disease . Br J Surg 2013 ; 100 : 1172 – 1179 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Reynolds MR , Apruzzese P, Galper BZ, et al. Cost-effectiveness of supervised exercise, stenting, and optimal medical care for claudication: results from the Claudication: Exercise Versus Endoluminal Revascularization (CLEVER) trial . J Am Heart Assoc 2014 ; 3 : e001233 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Gusi N , Reyes MC, Gonzalez-Guerrero JL, et al. Cost–utility of a walking programme for moderately depressed, obese, or overweight elderly women in primary care: A randomised controlled trial . BMC Public Health 2008 ; 8 : 231 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Barton GR , Sach TH, Jenkinson C, et al. Lifestyle interventions for knee pain in overweight and obese adults aged > or = 45: Economic evaluation of randomised controlled trial . BMJ 2009 ; 339 : b2273 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Neumann PJ , Cohen JT, Weinstein MC. Updating cost-effectiveness—the curious resilience of the $50,000-per-QALY threshold . N Engl J Med 2014 ; 371 : 796 – 797 . Google Scholar Crossref Search ADS PubMed WorldCat © The European Society of Cardiology 2020 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 2020

Dongiovanni, Paola; Ruscica, Massimiliano

doi: 10.1177/2047487319895402pmid: 31852301

Over the last 20 years, chronic lifestyle-related diseases such as visceral obesity, type 2 diabetes mellitus (T2DM), and non-alcoholic fatty liver disease (NAFLD) have exacerbated the burden of death and disability caused by cardiovascular diseases (CVDs). NAFLD is the leading cause of chronic liver disease worldwide, and it encompasses a wide spectrum of conditions, ranging from simple steatosis to non-alcoholic steatohepatitis (NASH), fibrosis, and hepatocellular carcinoma. Although NAFLD is associated with an increased risk of liver-related morbidity or mortality, it is now considered a multisystem disorder, which affects a variety of extra-hepatic organs, including the cardiovascular (CV) system.1 Patients with a more severe NAFLD diagnosis were associated with a higher risk of fatal (odds ratio (OR) 1.63, 95% confidence interval (CI) 1.06–2.48) and non-fatal (OR 2.52, 95% CI 1.52–4.18) CV events,2 although the epidemiological evidence is still not conclusive. The close connection between NAFLD and atherosclerotic CVD (ASCVD) was also the issue of a meta-analysis that evaluated cross-sectional and cohort studies reporting that NAFLD is associated with a raised prevalence (OR 1.81, 95% 1.23–2.66) and incidence (hazard ratio (HR) 1.37, 95% CI 1.10–1.72) of CVD, without any impact on CVD mortality (HR 1.10, 95% CI 0.86–1.41).3 In this context, Janssen et al.4 in their “In the news” article claimed that NAFLD patients should be screened for CVD on a regular basis. A discordant finding was instead drawn by Alexander et al.,5 who concluded that the risk of CVD should be assessed in the standard way in these patients as for the general population since NAFLD is not considered a risk enhancer. This analysis, which comprises more than 18 million European adults, reported that the diagnosis of NAFLD led to an HR for acute myocardial infarction of 1.01 (95% CI 0.91–1.12), after adjustment for systolic blood pressure, T2DM, total cholesterol level, statin use, and hypertension. Similar findings were found in the case of stroke: HR 1.18 (95% CI 1.11–1.24).5 This issue was the aim of ad hoc Mendelian randomization analysis which demonstrated that genetically high liver fat content is not the casual determinant of CVD risk (OR 0.98, 95% CI 0.94 -1.03), although an observational association of liver fat content and NAFLD with CVD is confirmed (OR 1.05, 95% CI 1.02–1.09).6 A validated CV risk scoring for NAFLD patients that is able to predict major CV events is still missing, although ASCVD remains a key cause of morbidity in NAFLD. At times, the application of the existing scores (Framingham Risk Score, SCORE, or QRisk2) has been fraught with perplexities, since they may underestimate the CV risk in these patients.7 A novel algorithm, generated by binary logistic regression including mean platelet volume, has been proposed as a NAFLD CV risk score.8 However, there is a need to assess a one-year risk score that allows high-risk patients to be on primary prophylaxis. It is not surprising that NAFLD associates with an increased risk of CVD, since they share common risk factors such as abdominal obesity, hypertension, atherogenic dyslipidemia, and insulin resistance/dysglycemia, all features encompassing the spectrum of metabolic syndrome.9 In addition to these classical risk factors, the authors listed multiple pathophysiological mechanisms such as systemic inflammation, endothelial dysfunction, oxidative stress, and plaque formation. Although a number of studies looking at markers of subclinical atherosclerosis found that coronary artery calcium score associates with NAFLD and the carotid intima media thickness (cIMT) increased proportionally with fatty liver index, cIMT progression has not been related to the risk of future CVD events.10 Relative to subclinical atherosclerosis, another indicator to be monitored is the quantitation of high-density lipoprotein (HDL) cholesterol efflux capacity (CEC), a better predictor of CVD risk compared to absolute plasma HDL-C level. HDL CEC was suppressed in NAFLD patients, suggesting that impaired HDL functions may be an independent risk factor for atherosclerosis in NAFLD.11 Hepatic fat content and the progression of liver damage have a strong heritable component. The common variants I148M in the Patatin-like phospholipase domain-containing protein 3 (PNPLA3) and E167K in the Transmembrane 6 superfamily member 2 (TM6SF2) genes predispose to NAFLD and advanced liver disease but protect against CVD.12,13 Both these variants, whose characterization is now strongly recommended in the management of NAFLD patients, have been associated with lower plasma lipid levels, thus explaining the negative correlation between these mutations and CVD. Our group recently demonstrated that the rs236918 variant in the proprotein subtilisin/kexin type 7 (PCSK7) gene is associated with dyslipidemia and more severe liver disease in NAFLD patients, thus representing an additional CV risk factor.14 In search of a biomarker able to predict CVD risk in NAFLD patients, the role of the PCSK9 merits consideration. So far described as one of the key regulators of low-density lipoprotein cholesterol, it is also involved in triglycerides metabolism and has been associated with raised hepatic fat accumulation.15,16 Among other determinants linking NAFLD to CVD, the authors proposed (i) inflammation, often raised in patients with NAFLD, which can increase endothelial dysfunction, alter vascular tone, and enhance vascular plaque formation; and (ii) the involvement of cardiomyopathy, valvular heart disease, and cardiac arrhythmias. All in all, although NAFLD has been associated with valvular aortic valve sclerosis and mitral annulus calcification, or with atrial fibrillation (adjusted OR 5.88; 95% CI 2.72–12.7), no clear causality can be inferred.17,18 In the absence of approved pharmacotherapy options for the treatment of NAFLD, it is crucial to modify those key steps involved in the prevention of NAFLD that overlap with those of cardiovascular prevention – that is, lifestyle intervention. According to guidelines, an energy deficit of 500–1000 kcal achieved by low–moderate fat, low-carbohydrate ketogenic, or high-protein diets and moderate-intensity aerobic exercise are required.19 Recently, Koutoukidis et al.20 have reported the central role of weight loss, which has to be behavioral-driven instead of pharmacological-centered (orlistat or sibutramine). Weight loss interventions have been associated with improvements in liver biomarkers (–9.81 Unit/Liter; 95% CI –13.12 to −6.50), histologically or radiologically measured liver steatosis (–1.48; 95% CI −2.27 to −0.70), histologic NAFLD activity score (–0.92; 95% CI −1.75 to −0.09), and the presence of NASH (OR 0.14; 95% CI 0.04–0.49).20 Finally, other possible bridges between NAFLD and the increased risk of CVD may be worth exploring: (i) gut microbiota (dysbiosis) being able to influence intestinal permeability to pro-inflammatory bacterial components,21 (ii) the innate or adaptative immune system, and (iii) mitochondrial dysfunction. Patients with NAFLD frequently develop CVD as these conditions share common metabolic risk factors. To stratify the risk, it is thus crucial to identify which NAFLD patients have to be screened for CVD, also considering the inherited genetic predisposition to the development and progression of NAFLD. Finally, relative to the pharmacological approaches, which are comparatively narrow, the candidate drugs under evaluation for NAFLD should be investigated in depth, possibly using CV endpoints. 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 Targher G , Day CP, Bonora E. Risk of cardiovascular disease in patients with nonalcoholic fatty liver disease . N Engl J Med 2010 ; 363 : 1341 – 1350 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Targher G , Byrne CD, Lonardo A, et al. Non-alcoholic fatty liver disease and risk of incident cardiovascular disease: a meta-analysis . J Hepatol 2016 ; 65 : 589 – 600 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Wu S , Wu F, Ding Y, et al. Association of non-alcoholic fatty liver disease with major adverse cardiovascular events: a systematic review and meta-analysis . Sci Rep 2016 ; 6 : 33386 – 33386 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Janssen A, Grobbee DE and Dendale P. Non-alcoholic fatty liver disease, a new and growing risk indicator for cardiovascular disease. Eur J Prev Cardiol. 2020; 27: 1059–1063 . 5 Alexander M , Loomis AK, van der Lei J, et al. Non-alcoholic fatty liver disease and risk of incident acute myocardial infarction and stroke: findings from matched cohort study of 18 million European adults . BMJ 2019 ; 367 : l5367 – l5367 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Lauridsen BK , Stender S, Kristensen TS, et al. Liver fat content, non-alcoholic fatty liver disease, and ischaemic heart disease: Mendelian randomization and meta-analysis of 279 013 individuals . Eur Heart J 2018 ; 39 : 385 – 393 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Stols-Goncalves D, Hovingh GK, Nieuwdorp M, et al. NAFLD and Atherosclerosis: two sides of the same dysmetabolic coin? Trends Endocrinol Metab 2019; 30: 891–902 . 8 Abeles RD , Mullish BH, Forlano R, et al. Derivation and validation of a cardiovascular risk score for prediction of major acute cardiovascular events in non-alcoholic fatty liver disease; the importance of an elevated mean platelet volume . Aliment Pharmacol Ther 2019 ; 49 : 1077 – 1085 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Angulo P . Nonalcoholic fatty liver disease . N Engl J Med 2002 ; 346 : 1221 – 1231 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Bahls M, Lorenz MW, Dorr M, et al. Progression of conventional cardiovascular risk factors and vascular disease risk in individuals: insights from the PROG-IMT consortium. Eur J Prev Cardiol. 2020; 27: 234–243 . 11 Fadaei R , Poustchi H, Meshkani R, et al. Impaired HDL cholesterol efflux capacity in patients with non-alcoholic fatty liver disease is associated with subclinical atherosclerosis . Sci Rep 2018 ; 8 : 11691 – 11691 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Dongiovanni P , Petta S, Maglio C, et al. Transmembrane 6 superfamily member 2 gene variant disentangles nonalcoholic steatohepatitis from cardiovascular disease . Hepatology 2015 ; 61 : 506 – 514 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Simons N , Isaacs A, Koek GH, et al. PNPLA3, TM6SF2, and MBOAT7 genotypes and coronary artery disease . Gastroenterology 2017 ; 152 : 912 – 913 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Dongiovanni P , Meroni M, Baselli G, et al. PCSK7 gene variation bridges atherogenic dyslipidemia with hepatic inflammation in NAFLD patients . J Lipid Res 2019 ; 60 : 1144 – 1153 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Macchi C , Banach M, Corsini A, et al. Changes in circulating pro-protein convertase subtilisin/kexin type 9 levels - experimental and clinical approaches with lipid-lowering agents . Eur J Prev Cardiol 2019 ; 26 : 930 – 949 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Ruscica M , Ferri N, Macchi C, et al. Liver fat accumulation is associated with circulating PCSK9 . Annals of Medicine 2016 ; 48 : 384 – 391 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Stahl EP , Dhindsa DS, Lee SK, et al. Nonalcoholic fatty liver disease and the heart: JACC state-of-the-art review . J Am Coll Cardiol 2019 ; 73 : 948 – 963 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Pastori D, Sciacqua A, Marcucci R, et al. Prevalence and impact of nonalcoholic fatty liver disease in atrial fibrillation. Mayo Clinic Proceedings. Epub ahead of print 18 November 2019. DOI: 10.1016/j.mayocp.2019.08.027 . 19 El-Agroudy NN , Kurzbach A, Rodionov RN, et al. Are lifestyle therapies effective for NAFLD treatment? Trends Endocrinol Metab 2019 ; 30 : 701 – 709 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Koutoukidis DA, Astbury NM, Tudor KE, et al. Association of weight loss interventions with changes in biomarkers of nonalcoholic fatty liver disease: a systematic review and meta-analysis. JAMA Intern Med. Epub ahead of print 1 July 2019. DOI: 10.1001/jamainternmed.2019.2248 . 21 Meroni M and Longo M. The role of probiotics in nonalcoholic fatty liver disease: a new insight into therapeutic strategies. Nutrients 2019; 11: 2642 . © The European Society of Cardiology 2020 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 2020

Janssen, Arne; Grobbee, Diederick E; Dendale, Paul

doi: 10.1177/2047487319891783pmid: 31801050

Introduction Non-alcoholic fatty liver disease (NAFLD) is a multisystem disease affecting extra-hepatic organs and regulatory pathways. It is characterised by the presence of ectopic fat in the liver which cannot be explained by alcoholic consumption.1 Non-alcoholic steatohepatitis (NASH) is an advanced form of NAFLD, where steatosis coexists with hepatocellular injury and inflammation. Out of 100 patients with NAFLD about 20 patients will develop NASH. When progressing even further it leads to hepatic necrosis, fibrosis, cirrhosis and even hepatocellular carcinoma.2,3 Is has been estimated that about one billion individuals worldwide have NAFLD with a prevalence of 20–30% in Western countries, making it the most frequent cause of liver disease in the Western world. These numbers are rapidly increasing in parallel with the global obesity pandemic.4 The progression of liver injury is a slow process, as is the case with atherosclerosis. Simple steatosis is considered to be a non-progressive condition. When present, NAFLD increases overall mortality by at least 35% and is associated with most risk factors of cardiovascular disease (CVD): obesity, type 2 diabetes mellitus, insulin resistance and the metabolic syndrome.4 In particular, (central) obesity is highly predictive of hepatic steatosis and disease progression, while in morbid obesity almost all patients present with steatosis and more than one-third have NASH.3 Research has shown that NAFLD is associated with CVD, particularly in patients with NASH. Not only do cardiovascular (CV) events occur more frequently, but also CV mortality is increased twofold as compared to the general population, making CVD the number one cause of death for NAFLD followed by malignancies and, only then, liver disease.5,6 Despite this association, NAFLD patients are currently not screened for CVD on a regular basis. More awareness is needed so that, ideally, each patient is referred upon (or even before) their diagnosis for CV risk factor screening and for treatment where needed. Association of NAFLD and CVD NAFLD is a risk factor for the development of CVD. When simple steatosis develops the CV risk already increases by 10–35%. When evolving into NASH and eventually cirrhosis, the CV risk increases even further by 12–40% for NASH and an extra 15% for cirrhosis.7 Research is still inconsistent as to whether this is independent of the presence of other risk factors. A recent meta-analysis using Mendelian randomization suggests a non-causal relationship and that the association is a result of confounding due to a number of common risk factors such as age, body mass index, smoking and hypertension, and may also be influenced by reverse causation. In this Mendelian randomization the genetic variant patatin-like phospholipase domain containing 3 (PNPLA3) was used as a proxy for liver fat content and was not shown to be associated with the risk of ischaemic heart disease.8,9 The increased risk of CVD may be caused by multiple pathophysiological mechanisms including systemic inflammation, endothelial dysfunction, systemic insulin resistance, oxidative stress, plaque formation and altered lipid metabolism.2,8 These mechanisms result in increased atherosclerosis, cardiomyopathy, arrhythmia and valvular heart disease, and ultimately increased CV mortality. Genetic factors also seem to be involved in the pathogenesis of CVD in patients with NAFLD.2,3,10 Atherosclerosis NAFLD has been shown to increase carotid intima media thickness and coronary artery calcification, with impaired flow-mediated vasodilation and increased carotid atherosclerotic plaques independent of metabolic syndrome characteristics. Patients are not only at risk of subclinical atherosclerosis but also require more frequent percutaneous coronary interventions with worse outcomes for patients with NAFLD who experience an acute coronary syndrome.2 Cardiomyopathy NAFLD is associated with an abnormal left ventricular (LV) structure and impaired diastolic function. The duration and severity of these abnormalities contributes to the increased risk of heart failure.6 More specifically, LV wall thickness and myocardial mass are greater in patients with NAFLD and echocardiography shows a lower early diastolic relaxation velocity, higher LV filling pressure and worse absolute global longitudinal strain.2 Valvular heart disease The presence of aortic valve sclerosis and mitral annulus calcification is also linked with NAFLD, independently of established CVD risk factors.1 Cardiac arrhythmias NAFLD seems to be linked with prolonged QTc interval, i.e. it is a powerful predictor of ventricular arrhythmias and sudden cardiac death,1 and there is growing evidence of an increased risk for atrial fibrillation. NAFLD has also been shown to be associated with autonomic dysfunction, which is another risk factor for arrhythmias.2 While other diseases are automatically referred for CVD screening, patients with NAFLD are only seen when CVD is already present. Figure 1 illustrates the diseases with a known association with CVD, where preventive cardiology should always be included. This contributes to the concept of lifelong CVD prevention, where preventive cardiology is considered in every person at every stage of life, starting at conception and ending at death (or even post-mortem). Only in this way can we truly reduce the burden of CVD. Figure 1. Open in new tabDownload slide Overview of diseases that have a known association with cardiovascular disease (CVD). Preventive cardiology is intertwined in these disciplines and should be included in the care paths of these diseases. This contributes to the concept of lifelong CVD prevention, where preventive cardiology is considered in every person in every stage of life, starting at conception and ending at death (or even post mortem). Even though the increased risk for CVD in non-alcoholic fatty liver disease (NAFLD) is clearly demonstrated, these patients are still not spotted on the preventive cardiology radar (Please refer to references17-31). Genetic factors Genetic factors play a key role in the susceptibility and progression of NAFLD. The most consistent genetic associations are seen with the PNPLA3 I148M gene and the transmembrane 6 superfamily member 2 (TM6SF2) E167K gene. Both of these genes are involved in lipid droplet remodelling and very low-density lipoprotein secretion. These findings suggest that hepatocellular accumulation of neutral lipids is harmful for the liver.11 This increase of hepatic fat is likely causally related to liver fibrosis, independent of inflammation, and also seems to modestly increase insulin resistance and the risk of type 2 diabetes mellitus.12 Other data also suggest an association with proprotein convertase subtilisin kexin type 9 (PCSK9), which is one of the key regulators of the low-density lipoprotein receptor. It could play a key role in the metabolism of triglyceride-rich lipoproteins. Circulating PCSK9 is correlated with the severity of steatosis, independently of metabolic confounders and liver damage.10,13 More research is needed in this area to specify the exact role of these (and other) genetic variants. Diagnosis The gold standard for the diagnosis of NAFLD is a liver biopsy, mainly for diagnosing NASH and staging fibrosis. However, diagnosis is often incidental on physical examination, through imaging or routine blood testing, accounting for approximately 7–11% of abnormal liver function tests.6 Clinical features that can be associated with NAFLD are obesity, metabolic syndrome, hypertension, insulin resistance, family history of NAFLD and hepatomegaly. Elevated levels of circulating biomarkers that should be considered are aspartate aminotransferase and alanine aminotransferase, cytokeratin 18 (CK-18) fragments, apolipoproten A1, total bilirubin, hyaluronic acid, C-reactive protein, fibroblast growth factor-21, interleukin 1 receptor antagonist, adiponectin and tumor necrosis factor alpha (TNF-α). However, at present, there is no readily available biomarker that reliably differentiates between simple steatosis and NASH. Regarding imaging techniques, the abdominal ultrasound is most commonly used, followed by magnetic resonance imaging and computed tomography.3,6 The role of CVD prevention in the treatment of NAFLD The prevention and treatment of NAFLD is inseparably linked with CVD prevention since common risk factors are shared. In primary prevention of NAFLD the main focus is on lifestyle modifications which include weight loss, adjusting nutrition, limiting alcohol consumption and increasing physical activity. In patients with confirmed NAFLD lifestyle, modifications are pivotal to achieve regression of fibrosis and resolution of steatohepatitis.2 A recent study has shown that a Mediterranean/low-carbohydrate diet decreases hepatic fat content even further in comparison with a low-fat diet, with a beneficial effect on cardiometabolic risk parameters.14 CV risk assessment can also play a role in primary prevention. It has been shown that Framingham Risk Score correlates with the degree of fibrosis in NAFLD patients. Risk scores should be routinely included in patients with NAFLD to stratify their risk.2 CV risk factors should be targeted individually. Statins should be considered in dyslipidaemia as previous studies suggest that statin therapy not only reduces CV morbidity but also improves liver enzyme test results.5 Statins also affect NAFLD itself, as research has shown an independent association between statin use and protection against steatosis, steatohepatitis (NASH) and fibrosis. This effect seems to be stronger when the PNPLA3 variant is not present.15 Currently, statins are under-prescribed because of an unjustified fear of hepatotoxicity. The CV as well as hepatic benefits seen with statin use appear to heavily outweigh the risk of hepatic toxicity.2 Similarly to statins, aspirin is thought to be effective against NAFLD by inhibiting the production of TNF-a and stimulating the expression of endothelial nitric oxide synthase and vascular endothelial growth factor, resulting in antioxidant activity. In hypertensive patients, angiotensin II receptor blockers (ARBs) have demonstrated a significant decrease in serum liver enzyme levels in several smaller trials. Well-designed randomised controlled trials are needed to confirm the effects of ARBs on NAFLD.2 Other treatments There is currently no accepted targeted treatment of NAFLD and NASH. Treatment is focused on targeting present risk factors and co-morbidities, as mentioned above.3 CV risk factors can be treated through use of statins, aspirin and anti-hypertensive drugs. Insulin resistance should be treated with insulin-sensitising agents, like glucagon-like peptide-1 analogues or metformin, improving insulin sensitivity and reducing hepatic gluconeogenesis.2 There is no approved specific treatment for the liver disease itself, although vitamin E could be used, with caution, in non-diabetic patients with NASH and fibrosis, but without cirrhosis and with no expected increased risk for prostate cancer.3 Note, however, that long-term use of this supplement has had no significant benefit in preventing major CV events.2 Bariatric surgery is another effective treatment for NAFLD, leading to a significant improvement in liver histology and even disappearance of NASH, and a reduction in fibrosis. Besides the effect on NASH, bariatric surgery also reduces CV risk factors. At present, bariatric surgery is only recommended for severely obese adolescents with significant steatohepatitis in whom therapeutic lifestyle intervention has been unsuccessful.2,3,6 Recently various pathogenic mechanisms are being tested for the treatment of NAFLD and NASH. This includes obeticholic acid (agonism of farnesoid X receptor), which decreases insulin sensitivity and hepatic gluconeogenesis but increases total serum cholesterol and low-density lipoprotein. Another potential mechanism is cenicriviroc, a chemokine receptor 2 and 5 antagonist, which promotes anti-inflammatory and antifibrotic effects in the liver. A recent study showed significant improvement in fibrosis without worsening NASH. Elafibranor, a dual peroxisome proliferator-activated receptor α/δ agonist, has also shown promising results in phase 2 trials by reducing fibrosis and improving liver transaminases and cardiometabolic parameters. There are many other mechanisms being investigated such as the inhibition of galectin-3 protein or the antagonism of toll-like receptors. Emerging data are promising and further updates from ongoing clinical trials are eagerly awaited.2,16 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 Byrne CD , Targher G. NAFLD: A multisystem disease . J Hepatol 2015 ; 62 : s47 – s64 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Stahl EP , Dhindsa DS, Lee SK, et al. Non-alcoholic fatty liver disease and the heart . J Am Coll Cardiol 2019 ; 73 : 948 – 963 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Machado M , Cortez-Pinto H. Non-alcoholic fatty liver disease: What the clinician needs to know . World J Gastroenterol 2014 ; 20 : 12956 – 12980 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Benedict M , Zhang X. Non-alcoholic fatty liver disease: An expanded review . World J Hepatol 2017 ; 9 : 715 – 732 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Motamed N , Rabiee B, Poustchi H, et al. Non-alcoholic fatty liver disease (NAFLD) and 10-year risk of cardiovascular diseases . Clin Res Hepatol Gastroenterol 2017 ; 41 : 31 – 38 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Temple JL , Cordero P, Li J, et al. A guide to non-alcoholic fatty liver disease in childhood and adolescence . Int J Mol Sci 2016 ; 17 : 947 – 947 . Google Scholar Crossref Search ADS WorldCat 7 Bhatia LS , Curzen NP, Calder PC, et al. Non-alcoholic fatty liver disease: A new and important cardiovascular risk factor? Eur Heart J 2012 ; 33 : 1190 – 1200 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Federico A , Dallio M, Masarone M, et al. The epidemiology of non-alcoholic fatty liver disease and its connection with cardiovascular disease: Role of endothelial dysfunction . Eur Rev Med Pharmacol Sci 2016 ; 20 : 4731 – 4741 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 9 Lauridsen BK , Stender S, Kristensen TS, et al. Liver fat content, non-alcoholic fatty liver disease, and ischaemic heart disease: Mendelian randomization and meta-analysis of 279 013 individuals . Eur Heart J 2018 ; 39 : 385 – 393 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Ruscica M , Ferri N, Macchi C, et al. Liver fat accumulation is associated with circulating PCSK9 . Ann Med 2016 ; 48 : 384 – 391 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Dongiovanni P , Romeo S, Valenti L. Genetic factors in the pathogenesis of nonalcoholic fatty liver and steatohepatitis . Biomed Res Int 2015 ; 1: 1–10 . Google Scholar OpenURL Placeholder Text WorldCat 12 Dongiovanni P , Stender S, Pietrelli A, et al. Causal relationship of hepatic fat with liver damage and insulin resistance in nonalcoholic fatty liver . J Intern Med 2018 ; 283 : 356 – 370 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Macchi C , Banach M, Corsini A, et al. Changes in circulating pro-protein convertase subtilisin/kexin type 9 levels – experimental and clinical approaches with lipid-lowering agents . Eur J Prev Cardiol 2019 ; 26 : 930 – 949 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Gepner Y , Shelef I, Komy O, et al. The beneficial effects of Mediterranean diet over low-fat diet may be mediated by decreasing hepatic fat content . J Hepatol 2019 ; 71 : 379 – 388 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Dongiovanni P , Petta S, Mannisto V, et al. Statin use and non-alcoholic steatohepatitis in at risk individuals . J Hepatol 2015 ; 63 : 705 – 712 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Dibba P , Li AA, Perumpail BJ, et al. Emerging therapeutic targets and experimental drugs for the treatment of NAFLD . Diseases 2018 ; 6 : 1 – 15 – 1–15 . Google Scholar OpenURL Placeholder Text WorldCat 17 Valmadrid CT, Klein R, Moss SE, et al. The risk of cardiovascular disease mortality associated with microalbuminuria and gross proteinuria in persons with older-onset diabetes mellitus. Arch Intern Med 2000; 160: 1093–1100 . 18 Keating NM, O'Malley AJ and Smith MR. Diabetes and cardiovascular disease during androgen deprivation therapy for prostate cancer. J Clin Oncol 2006; 24: 4448–4456 . 19 Lakshman R, Forouhi NG, Sharp SJ, et al. Early age at menarche associated with cardiovascular disease and mortality. J Clin Endocrinol Metab 2009; 94: 4953–4960 . 20 Bonnet F, Irving K, Terra JL, et al. Anxiety and depression are associated with unhealthy lifestyle in patients at risk of cardiovascular disease. Atherosclerosis 2005; 178: 339–344 . 21 Van Halm VP, Nurmohamed MT, Twisk JW, et al. Disease-modifying antirheumatic drugs are associated with a reduced risk for cardiovascular disease in patients with rheumatoid arthritis: A case control study. Arthritis Res Ther 2006; 8: 1–6 . 22 Bresalier RS, Sandler RS, Quan H, et al. Cardiovascular events associated with rofecoxib in a colorectal adenoma chemoprevention trial. N Eng J Med 2005; 352: 1092–1102 . 23 Stenvinkel P. Chronic kidney disease: A public health priority and harbinger of premature cardiovascular disease. J Intern Med 2010; 268: 456–467 . 24 Kvetny J, Heldgaard PE, Bladbjerg EM, et al. Subclinical hypothyroidism is associated with a low-grade inflammation, increased triglyceride levels and predicts cardiovascular disease in males below 50 years. Clin Endocrinol (Oxf) 2004; 61: 232–238 . 25 Hooi JD, Stoffers HEJH, Kester ADM, et al. Risk factors and cardiovascular diseases associated with asymptomatic peripheral arterial occlusive disease: The Limburg PAOD Study. Scand J Prim Health Care 1998; 16: 177–182 . 26 Schürks M, Rist PM, Bigal ME, et al. Migraine and cardiovascular disease: Systematic review and meta-analysis. BMJ 2009; 339: b3914 . 27 Islam F, Wu J, Jansson J, et al. Cardiovascular disease and HIV. HIV Med 2012; 13: 453–468 . 28 Gandaglia G, Briganti A, Jackson G, et al. A systematic review of the association between erectile dysfunction and cardiovascular disease. Eur Urol 2014; 65: 968–978 . 29 Gacci M, Corona G, Sebastianelli A, et al. Male lower urinary tract symptoms and cardiovascular events: A systematic review and meta-analysis. Eur Urol 2016; 70: 788–796 . 30 McClintic BR, McClintic JI, Bisognano JD, et al. The relationship between retinal microvascular abnormalities and coronary heart disease: A review. Am J Med 2010; 123: 374.e1–374.e7 . 31 De Hert M, Detraux J and Vancampfort D. The intriguing relationship between coronary heart disease and mental disorders. Dialogues Clin Neurosci 2018; 20: 31–40 . © The European Society of Cardiology 2020 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 2020

Boos, Christopher John

doi: 10.1177/2047487319894874pmid: 31840536

Hypertrophic cardiomyopathy (HCM) is the commonest inherited cardiomyopathy affecting approximately 0.2% of adults.1 It is defined as the presence of increased left ventricular wall thickness of 15 mm or greater in one or more left ventricular myocardial segments that is not explained solely by loading conditions.1 In one-third of cases there is evidence of resting left ventricular outflow tract (LVOT) obstruction. In another third LVOT obstruction is latent and is only manifested during manoeuvres (e.g. valsalva or exercise) that are known to affect left ventricular contraction and loading conditions. With increasing LVOT gradients (typically ≥50 mmHg) symptoms (e.g. syncope, breathlessness and chest pain) are far more likely to develop. When these are disabling despite medical treatment (e.g. beta-blockers, non-dihydropyridine calcium antagonists and disopyramide) septal myectomy (SM) with or without mitral valve surgery (11–20% of SM cases) or alcohol septal ablation (ASA) is usually recommended.1 The relative merits of both treatments are heavily influenced by patient age, comorbidities and the exact mechanisms of LVOT obstruction.1 The improvements in functional status and long-term mortality are generally similar for SM and ASA.2,3 However, peri-procedural mortality and stroke risk tend to be higher and the need for pacemaker implantation and re-intervention lower with SM.2,3 Several validated risk prediction models, utilising established HCM risk factors linked to adverse cardiovascular outcomes and sudden cardiac death (SCD), have been incorporated into clinical practice.1 They are designed to enhance clinical decision-making and the selection of patients for prophylactic implantable cardioverter defibrillator insertion for the prevention of SCD.4,5 Another increasingly recognised clinical risk stratification tool in HCM is that of cardiopulmonary exercise testing (CPET).6 One of its greatest assets in HCM management is in the objective assessment of functional exercise capacity and patient selection for SM or ASA treatment.6 In this issue of the journal, Smith et al. present the results of a single centre retrospective analysis of 295 (mean age 50 ± 14 years; 56% men) out of total of 1177 adult patients who underwent SM for HCM between 1996 and 2016.7 This represented all of the HCM patients in their SM cohort who had undergone a CPET both preceeding and within 6 months of their operation. Overall, 167 patients (56.6%) demonstrated an improvement in their peak oxygen consumption (VO2) and were classed as responders, with the remainder being non-responders. The independent predictors of non-response following SM were a higher pre-myectomy peak VO2, older age, dyslipidaemia, women (strongest predictor), a lack of cardiac rehabilitation enrolment and a lower body mass index (BMI). Furthermore, it was shown that non-response was independently associated with all-cause, but not cardiovascular, mortality over a mean follow-up of 11.3 years (adjusted hazard ratio 1.77, 95% confidence interval 1.06–3.34, P = 0.01). How do we interpret the findings of this interesting study?7 Many of the factors that have been linked to reduced exercise capacity and peak VO2 with HCM are also known to be related to increased mortality and adverse outcomes following SM. These include older age, female sex, significant comorbidities (e.g. anaemia, coronary artery disease and potentially dyslipidaemia) and higher levels of brain natriuretic peptides.6,8,9 Women with HCM tend to present at an older age, with worse symptoms and exercise tolerance and have higher LVOT gradients and mortality than men.10 It is therefore not surprising to discover that they were less likely to respond to SM than their male counterparts.7 The higher peak VO2 observed among the non-responders probably reflects the fact that these patients were less symptomatic and hence more unlikely to demonstrate an improvement in exercise capacity. The relationship between lower BMI and non-response is intriguing and more difficult to explain.7 In a recent observational study on 3282 HCM patients observed over a median follow-up of 6.8 years, Fumagalli et al. noted that obesity (31.7% of the cohort) was linked to an increased likelihood of obstructive physiology and adverse outcomes.10 Hence, it is reasonable to speculate in this current study by Smith et al. that the patients with the lowest BMI had less LVOT obstruction and higher peak VO2 and less to gain from SM. Conversely, the more obese patients, who one might naturally be less inclined to refer for SM, actually may have far more to gain from SM. 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 Elliott PM , Anastasakis A, Borger MA, et al. 2014 ESC Guidelines on diagnosis and management of hypertrophic cardiomyopathy: the Task Force for the Diagnosis and Management of Hypertrophic Cardiomyopathy of the European Society of Cardiology (ESC) . Eur Heart J 2014 ; 35 : 2733 – 2779 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Collis RA , Rahman MS, Watkinson O, et al. Outcomes following the surgical management of left ventricular outflow tract obstruction: a systematic review and meta-analysis . Int J Cardiol 2018 ; 265 : 62 – 70 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Osman M , Kheiri B, Osman K, et al. Alcohol septal ablation vs myectomy for symptomatic hypertrophic obstructive cardiomyopathy: systematic review and meta-analysis . Clin Cardiol 2019 ; 42 : 190 – 197 . Google Scholar Crossref Search ADS PubMed WorldCat 4 O’Mahony C , Jichi F, Pavlou M, et al. A novel clinical risk prediction model for sudden cardiac death in hypertrophic cardiomyopathy (HCM risk-SCD) . Eur Heart J 2014 ; 35 : 2010 c 2020 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Vriesendorp PA , Schinkel AF, Liebregts M, et al. Validation of the 2014 European Society of Cardiology guidelines risk prediction model for the primary prevention of sudden cardiac death in hypertrophic cardiomyopathy . Circ Arrhythmia Electrophysiol 2015 ; 8 : 829 – 835 . Google Scholar Crossref Search ADS WorldCat 6 Finocchiaro G , Haddad F, Knowles JW, et al. Cardiopulmonary responses and prognosis in hypertrophic cardiomyopathy: a potential role for comprehensive noninvasive hemodynamic assessment . JACC Heart Fail 2015 ; 3 : 408 – 418 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Smith JR , Layrissea V, Medina-Inojosaa JR, et al. Predictors of exercise capacity following septal myectomy in patients with hypertrophic cardiomyopathy . Eur J Prev Cardiol. 2020 ; 27 : 1066 – 1073 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Smith JR , Medina-Inojosa JR, Layrisse V, et al. Predictors of exercise capacity in patients with hypertrophic obstructive cardiomyopathy . J Clin Med 2018 ; 7 : 447 – 447 . Google Scholar Crossref Search ADS WorldCat 9 Panaich SS , Badheka AO, Chothani A, et al. Results of ventricular septal myectomy and hypertrophic cardiomyopathy (from Nationwide Inpatient Sample [1998–2010]) . Am J Cardiol 2014 ; 114 : 1390 – 1395 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Fumagalli C , Maurizi N, Day SM, et al. Association of obesity with adverse long-term outcomes in hypertrophic cardiomyopathy . JAMA Cardiol Epub ahead of print 6 November 2019. DOI: 10.1001/jamacardio.2019.4268. OpenURL Placeholder Text WorldCat © The European Society of Cardiology 2020 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 2020