OB E S I T Y COMORB I D I T Y Metabolically healthy overweight/obesity are associated with increased risk of cardiovascular disease in adults, even in the absence of metabolic risk factors: A systematic review and meta-analysis of prospective cohort studies Jacob Opio1 | Emma Croker2 | George S. Odongo3 | John Attia1,4 | Katie Wynne1,2,4 | Mark McEvoy1,4 1School of Medicine and Public Health, University of Newcastle, Callaghan, Australia 2Diabetes and Endocrinology, John Hunter Hospital, New Lambton Heights, Australia 3Department of Data Management and Statistics, Uganda Virus Research Institute, Entebbe, Uganda 4Centre for Clinical Epidemiology and Biostatistics, Level 4, Hunter Medical Research Institute, New Lambton Heights, Australia Correspondence Mark McEvoy, Centre for Clinical Epidemiology and Biostatistics, Level 4, The Hunter Medical Research Institute, The University of Newcastle, 1 Kookaburra Circuit, New Lambton Heights, NSW 2305, Australia. Email: mark.mcevoy@newcastle.edu.au Summary This review examined the risk of cardiovascular disease in adults with metabolically healthy overweight/obesity. A systematic review and meta-analysis using data from Medline, EMBASE, SCOPUS and Cochrane Library searched from inception up to 31st October 2019. We included prospective cohort studies of adults who are meta- bolically healthy or unhealthy. Outcomes were fatal and nonfatal cardiovascular events, all-cause mortality. Pooled relative risk was calculated for each outcome in populations with metabolically healthy overweight and metabolically healthy obesity using metabolically healthy normal weight as reference. A random-effects model was used for meta-analysis, and risk of bias assessment tool for nonrandomized studies assessed risk of bias within each study. Twenty-three prospective cohort studies with 4,492,723 participants were included. Cardiovascular disease risk was increased in metabolically healthy groups with overweight (RR = 1.34, CI: 1.23–1.46, n = 20, I2 = 90.3%) and obesity (RR = 1.58, CI: 1.34–1.85, n = 21, I2 = 92.2) compared with a reference group with metabolically healthy normal weight. Cardiovascular disease risk was similar irrespective of the number of risk factors used to define metabolically healthy and the risk remained in the group with no metabolic risk factors. Cardiovas- cular disease risk is increased in populations with overweight and obesity classified as metabolically healthy even when there were no metabolic risk factors. K E YWORD S cardiovascular disease, metabolic health, obesity, overweight 1 | INTRODUCTION The global prevalence of obesity has reached epidemic proportions with an estimated 671 million people having obesity in 2016.1 There is a well-established link between obesity, metabolic abnormalities (dyslipidaemia, dysglycaemia and hypertension) and cardiovascular disease (CVD),2 but it has been proposed that a subgroup of people with overweight/obesity and fewer or no metabolic abnormalities may not carry an increased risk of CVD.3 These subgroups are referred to as having metabolically healthy overweight (MHOW) and metabolically healthy obesity (MHO).4,5 This concept is complex, and the prevalence of MHO/OW varies between 10% and 40% depending on the type of population and the definition of metabolic health used in the study.3,6,7 The definition of the MHO/OW phenotype remains unclear and contentious.8 The number of metabolic parameters (tri- glycerides, high-density lipoprotein cholesterols, fasting blood glucose, Received: 5 June 2020 Revised: 7 July 2020 Accepted: 22 July 2020 DOI: 10.1111/obr.13127 Obesity Reviews. 2020;1–13. wileyonlinelibrary.com/journal/obr ©2020 World Obesity Federation 1 https://orcid.org/0000-0002-2235-9415 mailto:mark.mcevoy@newcastle.edu.au https://doi.org/10.1111/obr.13127 http://wileyonlinelibrary.com/journal/obr systolic and diastolic blood pressure) that need to be within normal parameters to be considered metabolically healthy has not been clearly defined. This uncertainty has led to inconsistent results from studies examining the association between MHO/OW by any defini- tion and development of CVD. The available systematic reviews and meta-analyses have con- cluded that MHO/OW is associated with an increased risk of develop- ing CVD.9–14 Kramer et al. combined the risk of all-cause mortality and cardiovascular events in their analysis,10 but some studies showed a stronger association for CVD mortality than all-cause mortality15,16 or no association between MHO/OW and all-cause mortality.12 Eckel et al. suggested that the cause of the high heterogeneity and con- flicting results observed in previous reviews was related to the absence of a standard definition for metabolic syndrome (MS) and the use of different outcome definitions.11 Since the review performed in 2016,12 there have been several large prospective cohort studies examining the relationship between MHO/OW and CVD with strict definitions of metabolic health (i.e., having no metabolic risk factors). The earlier systematic review by Eckel et al. pooled relative risk from three studies with strict definitions of metabolic health,11 and Zheng et al. reported pooled relative risk from two such studies.12 Both stud- ies reported no association between MHO phenotype with no meta- bolic risk factors and risk of CVD; however, these findings are unreliable because of the small number of studies. The latest system- atic review by Yeh et al. published early in 2019 failed to report on the risk of CVD in MHO participants who have no risk factors, included studies with an obesity cut-off less than 30 kg/m2 and exam- ined only four obesity phenotypes.14 Given the conflicting results from previous studies, the availability of more recent studies, includ- ing a total of seven studies of MHO/OW populations without meta- bolic risk factors, and the need to determine if overweight/obesity is associated with CVD in the absence of metabolic risk factors, we con- ducted a systematic review of the available literature to examine the association between MHO/OW and the risk of CVD events in adult men and women aged 18 years and older. 2 | METHODS AND MATERIAL This systematic review and meta-analysis were conducted according to the Meta-analysis of Observational Studies in Epidemiology guide- lines.17 The study protocol was registered in the International Pro- spective Register of Systematic Reviews18 registration number: CRD42019103850. 2.1 | Data source and searches A systematic literature search was performed using Medline, EMBASE, SCOPUS databases and Cochrane Library from inception up to the 31st October 2019. Each database was searched using the same comprehensive search terms (Table S1). Hand-search of refer- ences from relevant publications was also performed. Two independent reviewers analysed citations from the elec- tronic databases for relevant studies based on their title and abstract after removal of duplicate articles. Any disagreements were resolved by consensus or through consultation with a third investigator. 2.2 | Study selection Studies were considered eligible for inclusion if they met the following criteria: • prospective cohort design; • population included adult men and women 18 years and older without pre-existing CVD, diabetes mellitus, cancer or chronic kid- ney disease; • exposure consisted of a combination of body mass index (BMI) cat- egories or waist circumference and metabolic health and BMI cate- gories included: normal weight (BMI = 18.5–24.9 kg/m2), overweight (BMI = 25–29.9 kgm2) and obese (BMI ≥ 30kgm2); • metabolic health was defined in three separate ways as (i) absence of or (ii) presence of ≤1 or ≤2 metabolic abnormalities according to NCEP ATP III or IDF criteria of MS definition or (iii) other criteria (absence of hypertension, diabetes mellitus and dyslipidaemia); • reference group was individuals with metabolically healthy normal weight (MHNW); • exposure groups stratified according to metabolic health status and weight classification as follows: metabolically healthy normal weight (MHNW), metabolically healthy overweight (MHOW), metabolically healthy obese (MHO), metabolically unhealthy normal weight (MUNW), metabolically unhealthy overweight (MUOW) and metabolically unhealthy obesity (MUO); • included outcomes consisting of fatal and nonfatal cardiovascular events and all-cause mortality. Fatal and nonfatal CV events were defined as death due to CVD or one of the following: myocardial infarction, acute coronary syndrome, angina, coronary revasculari- zation, coronary artery bypass surgery, congestive heart failure, stroke and transient ischaemic attack; • if multiple studies were available from the same cohort, the most recent study was included. Studies were excluded using the following criteria: • subjects younger than 18 years; • subjects not human or with BMI in the underweight range; • studies with abstract or full text missing; • studies not stratified according to BMI categories and metabolic health; • cross-sectional studies, case studies, case reports, case series and letter to editors; • studies in languages other than English. 2 OPIO ET AL. MS components (risk factors) used to define metabolic health according to National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III) and International Diabetes Fed- eration (IDF): systolic BP, ≥130 mmHg, and diastolic BP, ≥85 mmHg; triglycerides, ≥1.7 mmol/L; high-density lipoprotein cholesterol (HDL cholesterol), ≤1.03 mmol/L (men) and ≤1.29 mmol/L (women); fasting glucose, ≥6.1 mmol/L (ATP III) and ≥5.6 mmol/L (IDF); waist circumference, ≥102 cm (men) and ≥88 cm (women) for ATP III and ≥94 cm (men) and ≥80 cm (women) for IDF. MS was defined as the presence of three or more of the above metabolic risk factors. 2.3 | Quality assessment Two independent reviewers assessed the quality of included cohort studies using the validated risk of bias assessment tool for non- randomized studies (RoBANS).19 The tool contains a table with six domains: participant selection, confounding variables, intervention measurement, blinding of outcome assessment, incomplete out- come data and selective outcome reporting. Each study was allo- cated low, high or unclear RoB in each domain. Studies with a low risk of bias for all six domains were considered high quality, studies with a low risk of bias for five domains were considered moderate quality and studies with a low risk of bias for four or less domains were considered low quality. The RoB scores are presented in (Table S2). 2.4 | Data extraction Two independent reviewers carried out data extraction using a stan- dardized form. The variables extracted are described inTable 1. 2.5 | Data analysis The pooled relative risk (RR) of CVD, all-cause mortality and car- diovascular mortality were calculated from the outcomes for the MHOW and MHO phenotypes with MHNW used as the reference population. Heterogeneity was investigated using Cochrane's Q test and I2 statistic. If heterogeneity was present (p value of Q test < 0.1 and I2 > 25%) a random-effects model was used. Sources of heterogeneity were explored by fitting each covariate (outcome definition, study quality, criteria for defining metabolic health, geographic location, duration of follow-up, number of risk factors in metabolically healthy and sample size) into a meta-regression model if these data were available. If the p value after adding the variable was less than 0.05, then that variable was considered a source of heterogeneity and sub- group analysis was performed accordingly. Publication bias was assessed using funnel plots and Egger's test. If the funnel plot was asymmetrical, a contour enhanced funnel plot was constructed to distinguish whether heterogeneity or publication bias was the cause of asymmetry. All analyses were conducted using STATA version 16. 3 | RESULTS 3.1 | Study characteristics There were 4,353 citations identified from the Medline, PubMed, EMBASE, SCOPUS and Cochrane Library databases. One hundred seven-eight studies were identified after screening abstracts and removing duplicates. Five studies were identified through manual searching, resulting in 183 studies receiving full-text review. Of the 183 studies, 23 prospective cohort studies with a total of 4,492,723 participants were eligible for systematic review and meta-analysis (Figure 1). Of the 23 studies analysed (Table 1), 18 studies included data for both male and female subjects6,7,16,20–34 with three studies providing data separated by sex.25,27,32 Two studies reported data for female participants,35,36 and three reported data for male subjects only.15,37,38 Participants were from diverse geographical locations: 11 studies were from Europe,7,16,20–22,25,30–33,38 six from North America,15,24,28,34–36 three from the Middle East,23,29,37 two from Asia26,27 and one from Oceania (Australia).6 No studies detailed the ethnic origins of their subjects. Sample size ranged from 1,65016 to 3.5 million subjects.20 Studies used different criteria to define metabolic health. Nine studies7,15,26–28,33,36–38 used the NCEP ATP III criteria,39 nine studies6,16,21–23,29–32 used IDF criteria40 and the remaining five stud- ies20,24,25,34,35 used absence of hypertension, diabetes and dyslipidaemia to define metabolic health. All studies used BMI = 25–<30 kgm2 and BMI ≥ 30 kg/m2 as the definition of overweight and obesity, respectively. Metabolic health was defined differently across studies: Nine used the absence of MS, three used the presence of two or less risk factors, four studies used one or less risk factors and seven studies used the absence of any risk factors. Seventeen studies reported CVD as the only outcome,6,7,20–30,32,35–37 four studies reported cardiovascular mortal- ity in addition to combined CVD and all-cause mortality as out- comes.15,16,31,38 Two studies reported all-cause mortality as the only outcome.33,34 Most studies6,7,15,16,20,22–24,27–29,31,35–38,41 reported CVD as a composite outcome that included the following presenta- tions of CVD: coronary artery disease (angina, ischaemic heart dis- ease and myocardial infarction), cerebrovascular disease or heart failure. RoB was evaluated using the RoBANS tool.19 All studies included in the meta-analysis achieved an overall quality score of 5–6/6. OPIO ET AL. 3 T A B L E 1 St ud y ch ar ac te ri st ic s A ut ho r Y ea r C o un tr y St ud y de si gn M ea n o r m ed ia n ye ar s o f fo llo w -u p Sa m pl e si ze % M al e M ea n ag e (y ea rs ) C ri te ri a fo r d ef in it io n o f m et ab o lic h ea lt h N u m b er o f m et ab o lic ri sk fa ct o rs in m et ab o lic al ly h ea lt h y O b es it y m ea su re m en t (c u t- o ff ) O u tc o m e A pp le to n et al . 2 0 1 3 A us tr al ia P ro sp ec ti ve co ho rt 8 .2 a 3 ,7 4 3 N o t re po rt ed N o t re po rt ed P re se nc e o f o n e o r le ss ri sk fa ct o rs ac co rd in g to ID F cr it er ia 0 o r 1 B M I≥ 3 0 kg /m 2 C V D A rn lo v et al . 2 0 0 9 Sw ed en P ro sp ec ti ve 3 0 .0 a 1 ,7 5 8 N o t re po rt ed N o t re po rt ed P re se nc e o f tw o o r le ss ri sk fa ct o rs ac co rd in g to A T P III cr it er ia ≤ 2 B M I≥ 3 0 kg /m 2 A C M ,C V M , M C E A un g et al . 2 0 1 4 U SA P ro sp ec ti ve st ud y 7 .4 a 3 ,7 0 0 N o t re po rt ed N o t re po rt ed P re se nc e o f tw o o r le ss m et ab o lic ab n o rm al it ie s ≤ 2 B M I≥ 3 0 kg /m 2 C V D B o et al . 2 0 1 2 It al y P ro sp ec ti ve co ho rt 9 .0 1 ,6 5 8 4 7 .0 N o t re po rt ed P re se nc e o f tw o o r le ss ri sk fa ct o rs ac co rd in g to ha rm o ni za ti o n de fi ni ti o n ≤ 2 B M I≥ 3 0 kg /m 2 C V D ,A C M , C V M C al ey ac he tt y et al . 2 0 1 7 U ni te d K in gd o m P ro sp ec ti ve co ho rt 5 .4 3 ,5 0 0 ,0 0 0 4 5 .5 N o t re po rt ed P re se nc e o f n o m et ab o lic ab no rm al it ie s (d ia be te s, hy pe rt en si o n o r hy pe rc ho le st er o la em ia ) 0 B M I≥ 3 0 kg /m 2 St ro ke ,C H D , H F D ha na et al . 2 0 1 6 T he N et he rl an ds P ro sp ec ti ve co ho rt 1 0 .3 a 5 ,3 1 4 4 0 .0 6 8 .0 P re se nc e o f tw o o r le ss ri sk fa ct o rs ac co rd in g to ID F cr it er ia ≤ 2 B M I≥ 3 0 kg /m 2 C V D E ck el et al . 2 0 1 8 U SA P ro sp ec ti ve co ho rt 2 4 .0 a 9 0 ,2 5 7 0 N o t re po rt ed P re se nc e o f n o m et ab o lic ab no rm al it ie s (d ia be te s, hy pe rt en si o n , hy pe rc ho le st er o la em ia ) 0 B M I≥ 3 0 kg /m 2 C V D ,M I, St ro ke . H an se n et al . 2 0 1 7 D en m ar k P ro sp ec ti ve 1 0 .6 6 ,2 3 8 4 8 .9 4 6 .3 m en 4 5 .7 w o m en P re se nc e o f n o m et ab o lic ri sk fa ct o rs 0 B M I≥ 3 0 kg /m 2 IH D H in no uh o 2 0 1 5 U ni te d K in gd o m P ro sp ec ti ve co ho rt 1 7 .4 7 ,1 2 2 6 9 .7 N o t re po rt ed P re se nc e o f tw o o r le ss ri sk fa ct o rs ac co rd in g to A T P III cr it er ia ≤ 2 B M I≥ 3 0 kg /m 2 C V D H o ss ei np an ah et al . 2 0 1 1 Ir an P ro sp ec ti ve 8 .4 a 6 ,2 1 5 4 3 .1 4 7 .4 P re se nc e o f n o dy sm et ab o lic st at us de fi n ed as h av in g m et ab o lic sy n d ro m e ac co rd in g to ID F cr it er ia o r di ab et es 0 o r 1 B M I≥ 3 0 kg /m 2 C V D H yu n- Ju ng Le e et al . 2 0 1 8 So ut h K o re a R et ro sp ec ti ve co ho rt 7 .4 3 5 4 ,0 8 3 5 2 .7 4 5 .8 P re se nc e o f n o ri sk fa ct o r ac co rd in g to A T P III cr it er ia 0 B M I≥ 3 0 kg /m 2 St ro ke K at zm ar zy k et al . 2 0 0 5 C an ad a O bs er va ti o na l co ho rt 1 0 .2 1 9 ,1 7 2 1 0 0 4 3 .1 P re se nc e o f tw o o r le ss ri sk fa ct o rs ac co rd in g to A T P III cr it er ia ≤ 2 B M I≥ 3 0 kg /m 2 A C M ,C V M 4 OPIO ET AL. T A B L E 1 (C o nt in ue d) A ut ho r Y ea r C o un tr y St ud y de si gn M ea n o r m ed ia n ye ar s o f fo llo w -u p Sa m pl e si ze % M al e M ea n ag e (y ea rs ) C ri te ri a fo r d ef in it io n o f m et ab o lic h ea lt h N u m b er o f m et ab o lic ri sk fa ct o rs in m et ab o lic al ly h ea lt h y O b es it y m ea su re m en t (c u t- o ff ) O u tc o m e La ss al e et al . 2 0 1 7 D en m ar k, F ra nc e, G er m an y, G re ec e, It al y, T he N et he rl an ds , N o rw ay ,S pa in , Sw ed en , U ni te d K in gd o m P ro sp ec ti ve ca se -c o ho rt 1 2 .2 a 1 0 ,4 7 4 3 7 .0 5 3 .6 P re se nc e o f tw o o r le ss ri sk fa ct o rs ac co rd in g to ID F cr it er ia ≤ 2 B M I≥ 3 0 kg /m 2 C H D Lo pr in zi et al . 2 0 1 7 U SA P ro sp ec ti ve 8 .6 a 7 ,5 7 9 4 9 .0 4 6 .0 P re se nc e o f n o ri sk fa ct o r ac co rd in g to th e A m er ic an N at io na lH ea rt ,L u n g, an d B lo o d In st it u te G ui de lin es 0 B M I≥ 3 0 kg /m 2 A C M Lu et al . 2 0 1 4 C hi na P ro sp ec ti ve o bs er va ti o na l 1 0 .0 2 1 1 ,6 4 1 3 4 .1 N o t re po rt ed P re se nc e o f tw o o r le ss ri sk fa ct o rs ac co rd in g to A T P III cr it er ia ≤ 2 B M I≥ 3 0 kg /m 2 C V D M ei gs et al . 2 0 0 6 U SA P ro sp ec ti ve o bs er va ti o na l 1 1 .4 2 ,9 0 2 4 5 .0 5 3 .0 P re se nc e o f tw o o r le ss ri sk fa ct o r ac co rd in g to A T P III cr it er ia ≤ 2 B M I≥ 3 0 kg /m 2 C V D M ir za ei et al . 2 0 1 7 Ir an P ro sp ec ti ve co ho rt 1 2 .0 7 ,1 6 7 4 4 .8 4 6 .3 P re se nc e o f tw o o r le ss ri sk fa ct o rs ac co rd in g to ID F cr it er ia ≤ 2 B M I≥ 3 0 kg /m 2 C V D M o rk ed al et al . 2 0 1 4 N o rw ay P ro sp ec ti ve co ho rt 1 2 .2 6 1 ,2 9 9 4 6 .1 N o t re po rt ed P re se nc e o f o n e o r n o ri sk fa ct o rs ac co rd in g to ID F cr it er ia 0 o r 1 B M I≥ 3 0 kg /m 2 o r W ai st ci rc u m fe re n ce > 9 4 cm in m en > 8 0 cm in w o m en A M I, H F O rt eg a et al . 2 0 1 3 Sw ed en Lo ng it ud in al 2 4 .0 4 3 ,2 6 5 7 5 .7 4 4 .2 P re se nc e o f o n e o r n o ri sk fa ct o r ac co rd in g to ID F cr it er ia 0 o r 1 B M I≥ 3 0 kg /m 2 A C M ,C V D , C V M So ng et al . 2 0 0 7 U SA P ro sp ec ti ve 1 0 .2 2 5 ,6 2 6 0 N o t re po rt ed P re se nc e o f tw o o r le ss ri sk fa ct o rs ac co rd in g to A T P III cr it er ia ≤ 2 B M I≥ 3 0 kg /m 2 C V D ,C H D , St ro ke T ho m se n et al . 2 0 1 4 D en m ar k P ro sp ec ti ve 3 .6 a 7 1 ,5 2 7 4 3 .7 N o t re po rt ed P re se nc e o f tw o o r le ss ri sk fa ct o rs ac co rd in g to ID F cr it er ia ≤ 2 B M I≥ 3 0 kg /m 2 M I, IH D (C o n ti n u es ) OPIO ET AL. 5 3.2 | MHOW and risk of cardiovascular events As one study31 did not report raw data for MHOW, data were pooled from 20 studies, which investigated the risk of CVD in those with MHOW compared with MHNW (Table S3). There was an increased risk of CVD in the group with MHOW compared with the group with MHNW (RR = 1.34, CI: 1.23–1.46, n = 20, I2 = 90.3%); however, heterogeneity was high (Figure 2). The risk of CVD did not differ significantly from the overall risk regardless of whether the CVD outcome was defined as composite (RR = 1.36, CI: 1.20–1.53, n = 15, I2 = 91.8%) or single entities of CVD (noncomposite) (RR = 1.30, CI: 1.18–1.43, n = 5, I2 = 62.8%). Relative risk did not differ according to study quality (high quality: RR = 1.43, CI: 1.29–1.58, n = 11, I2 = 80.1%; moderate quality: RR = 1.22, CI: 1.07–1.38, n = 9, I2 = 82.5%); sample size (≥6,000: RR = 1.33, CI: 1.19–1.47, n = 14, I2 = 93.6%; <6,000: RR = 1.36, CI: 1.14–1.64, n = 6, I2 = 50.7%); geographic location (Europe: RR = 1.34, CI: 1.18–1.52, n = 9, I2 = 93.7%; North America: RR = 1.54, CI: 1.18–2.03, n = 5, I2 = 78.3%); or study follow-up duration (<10 years: RR = 1.34, CI: 1.08–1.68, n = 7, I2 = 86.8%; ≥10 years: RR = 1.34, CI: 1.22–1.47, n = 13, I2 = 82.6%). Interest- ingly, the relative risk was similar irrespective of the quantity of risk factors used to define metabolically healthy (0: RR = 1.51, CI: 1.21–1.88, n = 5, I2 = 86.2%; ≤1: RR = 1.25, CI: 0.87–1.80, n = 3, I2 = 74.8%; ≤2: RR = 1.29, CI: 1.21–1.38, n = 12, I2 = 54.5%). The risk of CVD remained high even when there were no metabolic risk factors. The risk of CVD was different if the study used the ATP III or IDF criteria when compared with those studies that used ‘absence of hypertension, diabetes and dyslipidaemia’ (‘other’ criteria) to define metabolic health (NCEP ATP III: RR = 1.31, CI: 1.20–1.44, n = 8, I2 = 44.6%; IDF: RR = 1.22, CI: 1.07–1.39, n = 8, I2 = 74.5%; ‘Other’ criteria: RR = 1.68, CI: 1.53–1.85, n = 4, I2 = 33.5%). There was a statistically significant subgroup difference for criteria used to define metabolic health (Q = 20.15, p < 0.01) (Figure 2). Furthermore, the pooled effect size for risk of CVD in MHOW compared with MHNW did not change when the study by Caleyachetty et al. (2017) with the greatest weighting was removed from analysis (data not shown). However, these results should be interpreted with caution because of the small number of studies used in the subgroup analysis. The pooled relative risk from studies with disaggregated data for females and males was different (females: RR = 1.46, CI: 1.26–1.69, n = 3, I2 = 44.3%; males: RR = 1.22, CI: 1.14–1.30, n = 3, I2 = 0.0%). There was a statistically significant subgroup dif- ference for gender (Q = 4.75, p = 0.03) (Figure S1). However, cau- tion is needed in the interpretation because of the small number of studies. There was a significantly increased risk of all-cause mortality (RR = 1.22, CI: 1.02–1.46, n = 5, I2 = 48.8%) and CVD mortality (RR = 1.34, CI: 1.12–1.61 n = 4, I2 = 0.0%) in the group with MHOW compared with the group with MHNW (Figures S2 and S3); however, there were too few studies to reliably assess this risk.T A B L E 1 (C o nt in ue d) A ut ho r Y ea r C o un tr y St ud y de si gn M ea n o r m ed ia n ye ar s o f fo llo w -u p Sa m pl e si ze % M al e M ea n ag e (y ea rs ) C ri te ri a fo r d ef in it io n o f m et ab o lic h ea lt h N u m b er o f m et ab o lic ri sk fa ct o rs in m et ab o lic al ly h ea lt h y O b es it y m ea su re m en t (c u t- o ff ) O u tc o m e T w ig et al . 2 0 1 5 Is ra el C o ho rt 6 .1 3 1 ,6 8 4 1 0 0 3 1 .2 P re se nc e o f n o n e o f th e m et ab o lic ab n o rm al it ie s ac co rd in g to A T P III cr it er ia 0 B M I≥ 3 0 kg /m 2 C H D va n de r A 2 0 1 4 T he N et he rl an ds P ro sp ec ti ve co ho rt 1 3 .4 2 0 ,2 9 9 4 5 .8 N o t re po rt ed P re se nc e o f n o ri sk fa ct o rs ac co rd in g to A T P III cr it er ia 0 W ai st ci rc u m fe re n ce ≥ 1 0 2 cm in m en ≥ 8 8 cm in w o m en A C M A bb re vi at io ns :A C M ,a ll- ca us e m o rt al it y; A T P III ,A du lt T re at m en t P an el III ;C A D ,c o ro na ry ar te ry di se as e; C H D ,c o ro na ry he ar t di se as e; C V D ,c ar di o va sc u la r d is ea se ;C V M ,c ar d io va sc u la r m o rt al it y; ID F ,I n te rn a- ti o na lD ia be te s F ed er at io n; IH D :i sc ha em ic he ar t di se as e; M I, m yo ca rd ia li nf ar ct io n. a M ed ia n ag e. 6 OPIO ET AL. 3.3 | MHO and risk of CVD Twenty-one studies investigated the association between MHO and risk of CVD (Table S3). There was an increased risk of CVD in the group with MHO compared with the group with MHNW (RR = 1.58, CI: 1.34–1.85; I2 = 92.2%); however, heterogeneity was high (Figure 2). This risk of CVD did not differ significantly from the overall effect size regardless of whether the CVD out- come was defined as composite (RR = 1.64, CI: 1.37–1.96, n = 16, I2 = 88.4%) or single entities of CVD (noncomposite) (RR = 1.43, CI: 0.99–2.07, n = 5, I2 = 91.8%). The pooled RR did not differ sig- nificantly according to study quality (high: RR = 1.58, CI: 1.38–1.81, n = 12, I2 = 69.8%; moderate: RR = 1.67, CI: 1.07–2.62, n = 9, I2 = 96.4%); sample size (≥6,000: RR = 1.57, CI: 1.31–1.88, n = 15, I2 = 93.9%; <6,000: RR = 1.65, CI: 1.12–2.42, n = 6, I2 = 67.8%); geographic location (Europe: RR = 1.47, CI: 1.22–1.76, n = 10, I2 = 90.8%; North America: RR = 1.95, CI: 1.36–2.80, n = 5, I2 = 78.0%); or study follow-up duration (<10 years: RR = 1.89, CI: 1.13–3.17, n = 7, I2 = 94.5%; ≥10 years: RR = 1.51, CI: 1.27–1.79, n = 14, I2 = 83.8%). As for the MHOW group, the risk in the MHO group was similarly increased irrespective of the quantity of risk factors used to define metaboli- cally healthy (0: RR = 2.18, CI: 1.28–3.71, n = 5, I2 = 94.7%; ≤1: RR = 1.29, CI: 1.10–1.52, n = 4, I2 = 0.0%; ≤2: RR = 1.54, CI: 1.26–1.88, n = 12, I2 = 86.6%). The risk of CVD remained high even when there were no metabolic risk factors. The risk of CVD differed according to whether ATP III or IDF criteria were used, compared with those studies that used absence of hypertension, diabetes and dyslipidaemia (‘other’ criteria) to define metabolic health (NCEP ATP III: RR = 1.68, CI: 1.36–2.08, n = 8, I2 = 63.2%; IDF: RR = 1.25, CI: 1.04–1.51, n = 9, I2 = 70.9%; other criteria: RR = 2.32, CI: 1.55–3.49, n = 4, I2 = 91.8%). There was a statistically significant subgroup difference for criteria used to define metabolic health (Q = 9.23, p = 0.01) (Figure 2). Furthermore, the pooled effect size for risk of CVD in MHO compared with MHNW did not change when the study by Caleyachetty et al. (2017) with the greatest weighting was removed from analysis (data not shown). Again, these results should be interpreted with caution because of the small number of studies. The pooled RR from studies with disaggregated data for females and males were similar (females: RR = 1.71, CI: 1.28–2.28, n = 3, I2 = 61.1%; males: RR = 2.15, CI: 1.88–2.46, n = 3, I2 = 0.0%). Although F IGURE 1 Flow chart for literature search, study selection and reasons for exclusion OPIO ET AL. 7 the risk of CVD appears to be greater in males than females, the num- ber of studies was too few to reliably comment on the difference. There was a significantly increased risk of all-cause mortality (RR = 1.59, CI: 1.02–2.47, n = 6, I2 = 86.2%) but not of F IGURE 2 A, Meta-analysis of risk of cardiovascular events in participants with metabolically healthy overweight compared with metabolically healthy normal weight participants. B, Meta-analysis of risk of cardiovascular events in participants with metabolically healthy obesity compared with metabolically healthy normal weight participants. C, Meta-analysis of risk of cardiovascular events in participants with metabolically healthy overweight compared with metabolically healthy normal weight participants by criteria for defining metabolic health. D, Meta-analysis of risk of cardiovascular events in participants with metabolically healthy obesity compared with metabolically healthy normal weight participants by criteria for defining metabolic health 8 OPIO ET AL. cardiovascular mortality (RR = 2.22, CI: 0.96–5.11, n = 4, I2 = 81.9) in the group with MHO compared with the group with MHNW (Figures S4 and S5). It is important to note that although overweight/obesity increases the risk of CVD in individuals who are metabolically-healthy, this study also determined that metabolic abnormality confers an even greater risk of CVD in individuals with normal weight (pooled RR = 3.07; CI: 2.27–4.15, n = 20, I2 = 99.17%) (data not shown). 3.4 | Investigating heterogeneity and publication bias The results of the meta-analysis for MHOW had high heterogeneity except for studies with a sample size <6,000 (I2 = 50.68%) and those studies that used ATP III criteria to define metabolic health (I2 = 44.62%). High heterogeneity was also observed in the MHO analysis except in studies where metabolic health was defined by ≤1 risk factor (I2 = 0.00%) and the high-quality studies where the hetero- geneity was moderate (I2 = 69.81%). Despite the high heterogeneity, the overall effect size did not significantly differ across subgroups in both the MHOW and MHO analyses; meta-regression and analysis of subgroup difference identified that the criteria used to define meta- bolic health in both MHOW and MHO populations, and gender in MHOW populations as a cause of heterogeneity (Tables 2 and 3). The MHOW analysis showed no evidence of publication bias using a funnel plot and contour-enhanced funnel plot (Figures S6 and S7) or Egger's test (Table 2). The MHO analysis showed no evidence of publication bias using a funnel plot and contour-enhanced funnel plot (Figures S8 and S9) or Egger's test, except within studies grouped according to the criteria used to define metabolic health in which a small study effect was observed for studies using ‘absence of hyper- tension, diabetes and dyslipidaemia’ to define metabolic health (Egger's test Coef = 4.67; p value = 0.001) and studies with sample size <6,000 (Egger's test Coef = 3.19; p value = 0.001) (Table 3). 4 | DISCUSSION This systematic review is the most up to date meta-analysis per- formed in the field with obesity categorized into six phenotypes and includes nine new studies with an additional 4,012,796 participants.20–22,25,26,29,34,35,37 Some previous reviews included stud- ies which classified obesity into two categories by either combining normal weight and overweight or using obesity cut-off of 25 kg/m2. Unlike previous reviews this review did not include individuals with overweight in the comparator population.9,14 The use of a comparator group with MHNW enabled a more accurate estimate of the risk of CVD, cardiovascular mortality and all-cause mortality attributable to having MHOW/MHO. The data confirmed an increased risk of CVD in people classified as having MHOW or MHO when compared with a cohort with normal weight and metabolically healthy, consistent with literature by Kramer et al.,10 Fan et al.,9 Eckel et al.,11 Zheng et al.,12 Ortega et al.13 and Yeh et al.14 However, this data also identified an increased risk of CVD in the population with MHOW/MHO in the absence of any met- abolic risk factors, perhaps the most robust definition of optimal met- abolic health. This could be interpreted as evidence that the concept of MHOW/obesity does not exist. This is a new finding, and previous studies by Eckel11 and Zheng12 did not identify a statistically signifi- cant increased risk of CVD in populations with MHOW and MHO without metabolic risk factors. This finding may have practical implica- tions where individuals with overweight or obesity without established CVD should be encouraged to achieve normal weight to decrease the risk of developing CVD even in the absence of cardio- metabolic risk factors. However, caution is needed when interpreting these new data because of the high heterogeneity and limited numbers of studies. There was no clear dose–response curve established, that is, no graded increase in CVD risk when 0, ≤1 or ≤2 MS risk factor criteria were used for defining metabolically healthy within the groups with MHOW or MHO. However, the limited power from the small number of studies in each group could have resulted in a failure to detect a difference in effect size. Many studies reported the outcome data for participants without risk factors together with those having one or two risk factors in the definition of metabolic health, making it a chal- lenge to relate the magnitude of cardiovascular risk directly to the number of risk factors with certainty. This study revealed evidence of an increased risk of cardiovascu- lar mortality and all-cause mortality in people with MHOW/MHO. We performed meta-regressions and subgroup analyses to investi- gate heterogeneity and demonstrated that the risk of CVD in groups with MHO and MHOW compared with MHNW was not affected by the outcome definition, sample size, duration of follow-up, geographic location or quantity of metabolic risk factors used in the definition. The pooled effect size for risk of CVD in groups with MHOW/MHO com- pared with MHNW did not change when the study by Caleyachetty et al. (2017) with the greatest weighting was removed from analysis. All included studies were of high (n = 11) or moderate quality (n = 10) and there was no indication that quality affected the overall effect sizes; however, there was less heterogeneity in high-quality studies examin- ing the risk of CVD in populations with MHO (I2 = 69.81%). The results suggest that males and females with MHOW or MHO are at increased risk of CVD. The risk appeared to be greater for MHOW in females than males but was greater for MHO in males than females; this raises the possibility of differential effects between men and women, but this would need to be confirmed with much larger samples. A limitation of these findings is the high heterogeneity observed for each pooled relative risk. Meta-regression analysis in both groups with MHO and MHOW identified that the criteria used by each study for defining metabolic health were a cause of some het- erogeneity. There was a noticeable difference in the pooled relative risk according to whether ATP III or IDF criteria were used to define metabolic health compared with those studies that used absence of hypertension, diabetes and dyslipidaemia to define metabolic health. This emphasizes the need for a standard criteria that can be used to define metabolic health in future research. Another limitation was that most studies did not include cardiorespiratory fitness (CRF) and OPIO ET AL. 9 physical activity (PA). The review by Ortega et al. found that the difference in all-cause mortality and CVD mortality/morbidity between MHO and MHNW was largely explained by difference in CRF between the two phenotypes.13 A state-of-the-art-review on prevention and treatment of obesity indicated the significance of including CRF and PA as covariates in the models when comparing risk of CVD in MHO and MHNW.42 Moholdt T et al. showed that in coronary artery disease, PA was more important than weight for predicting survival43 and changes in PA were much more important than changes in weight for predicting survival.44 5 | RECOMMENDATIONS This research suggests that there is increased risk of CVD in people with MHOW or MHO. There is a need for a standard definition of metabolic health. It is recommended that future research examining the relationship between MHOW, MHO and the risk of CVD: • categorize metabolic risk factors into discrete groups of no risk fac- tors, one, two or three or more risk factors and define metaboli- cally healthy as having no risk factors. TABLE 2 Subgroups analysis of the risk of cardiovascular disease in participants with metabolically healthy overweight compared with participants with metabolically healthy normal weight Subgroup Effect size RR (95% CI) Number of studies Egger's test for small study effects: Statistic & p value Test of group difference: Statistic & p value Overall 1.34 (1.23–1.46) 20 z = −0.22 _ p = 0.83 Outcome Composite 1.36 (1.20–1.53) 15 z = −0.05, p = 0.96 Qb = 0.32 Noncomposite 1.30 (1.18–1.43) 5 z = −0.19, p = 0.85 p = 0.57 Quality High 1.43 (1.29–1.58) 11 z = −0.58, p = 0.56 Qb = 3.72 Moderate 1.22 (1.07–1.38) 9 z = 0.60, p = 0.55 p = 0.05 Sample size <6,000 1.36 (1.14–1.64) 6 z = 1.55, p = 0.12 Qb = 0.07 ≥6,000 1.33 (1.19–1.47) 14 z = −1.49, p = 0.14 p = 0.80 Criteria for defining metabolic health ATP III 1.31 (1.20–1.44) 8 z = 0.45, p = 0.66 Qb = 20.15 IDF 1.22 (1.07–1.39) 8 z = −1.23, p = 0.22 P < 0.001 Other 1.68 (1.53–1.85) 4 z = 0.90, p = 0.37 Number of risk factors in metabolically healthy 0 1.51(1.21–1.88) 5 z = −0.00, p = 0.99 Qb = 1.87 1 or less 1.25 (0.87–1.80) 3 z = −0.19, p = 0.85 p = 0.39 2 or less 1.29 (1.21–1.38) 12 z = 0.00, p = 0.99 Geographic location Asia 1.27 (1.00–1.63) 2 Qb = 3.92 Australia 1.53 (0.92–2.53) 1 p = 0.42 Europe 1.35 (1.23–1.48) 9 Europe: z = −0.99, p = 0.32 Middle East 0.99 (0.67–1.46) 3 North America 1.54 (1.20–1.96) 5 North America: z = 0.82, p = 0.41 Duration of follow up <10 years 1.34 (1.08–1.68) 7 z = 0.62, p = 0.5360 Qb = 0.00 10+ years 1.34 (1.22–1.47) 13 z = −0.53, p = 0.5947 p = 0.98 Gender Female 1.46 (1.26–1.69) 3 p = 0.94 Qb = 4.75 Male 1.22 (1.14–1.30) 3 p = 0.46 p = 0.03 Note: Egger's test of small study effects and meta-regression analysis were not done for geographic location because of small numbers in some groups. Gender: studies with mixed females and males participants. z Test statistics for Egger's test. Qb Test statistics for test of group difference. p = p Value. ATP III: Adult Treatment Panel III criteria for metabolic syndrome. IDF: International Diabetes Federation criteria for metabolic syndrome. Other: metabolic health defined by absence of hypertension, diabetes and dyslipidaemia. 10 OPIO ET AL. • design prospective cohorts with sufficient sample size and follow- up duration to determine CVD risk within population subgroups. • consider inclusion of other factors associated with metabolic health (e.g., CRF, PA, hepatic fat and inflammatory status deter- mined by C-reactive protein levels) • report changes in metabolic parameters over time. • include and report data on participants from other regions (e.g., Oceania, Africa, South America, Asia and Middle East) to increase diversity of the study population. 6 | CONCLUSIONS This systematic review and meta-analysis found an increased risk of CVD in populations with overweight and obesity considered metabolically healthy compared with those classified as normal weight and metabolically healthy. The risk was similar regardless of whether the definition of metabolically healthy included those who had none, one or less, two or less metabolic risk factors. The risk of CVD remained high even when there were no metabolic TABLE 3 Subgroups analysis of the risk of cardiovascular disease in participants with metabolically healthy obesity compared with participants with metabolically healthy normal weight Subgroup Effect size (95% CI) Number of studies Egger's test for small study effects: Statistic & p value Test of group difference: Statistic & p value Overall 1.58 (1.34–1.85) 21 z = 1.46 - p = 0.14 Outcome Composite 1.64 (1.37–1.96) 16 z = 1.12, p = 0.26 Qb = 0.39 Noncomposite 1.43 (0.99–2.07) 5 z = 0.58, p = 0.56 p = 0.53 Quality High 1.58 (1.38–1.81) 12 z = 0.32, p = 0.75 Qb = 0.05 Moderate 1.67 (1.07–2.62) 9 z = 1.10, p = 0.27 p = 0.82 Sample size <6,000 1.65 (1.12–2.42) 6 z = 0.78, p = 0.44 Qb = 0.05 ≥6,000 1.57 (1.31–1.88) 15 z = 3.19, p = 0.001 p = 0.82 Criteria for defining metabolic health ATP III 1.68 (1.36–2.08) 8 z = 0.92, p = 0.36 Qb = 9.23 IDF 1.25 (1.04–1.51) 9 z = 0.26, p = 0.80 p = 0.01 Other 2.32 (1.55–3.49) 4 z = 4.67, p < 0.001 Number of risk factors in metabolically healthy 0 2.18(1.28–3.71) 5 z = o.75, p = 0.45 Qb = 4.46 1 or less 1.29 (1.10–1.52) 4 z = −0.76, p = 0.45 p = 0.11 2 or less 1.54 (1.26–1.88) 12 z = 1.48, p = 0.14 Geographic location Asia 1.47 (0.71–3.05) 2 Australia 1.36 (0.73–2.54) 1 Qb = 2.16 Europe 1.46 (1.23–1.74) 10 Europe: z = 1.76, p = 0.08 p = 0.71 Middle East 1.67 (0.46–5.98) 3 North America 1.95 (1.36–2.78) 5 North America: z = 1.23, p = 0.22 Duration of follow up <10 years 1.89 (1.13–3.17) 7 z = 0.9, p = 0.37 Qb = 0.66 10+ years 1.51 (1.27–1.79) 14 z = 0.68, p = 0.49 p = 0.42 Gender Female 1.71 (1.28–2.29) 3 p = 0.51 Qb = 1.99 Male 2.15 (1.88–2.46) 3 p = 0.18 p = 0.16 Note: Egger's test of small study effects and meta-regression analysis were not done for geographic location because of small numbers in some groups. Gender: studies with mixed females and males participants. z Statistic for Egger's tests. Qb Statistics for test of group difference. p = p Value. ATP III: Adult Treatment Panel III criteria for metabolic syndrome. IDF: International Diabetes Federation criteria for metabolic syndrome. Other: Metabolic health defined by absence of hypertension, diabetes and dyslipidaemia. OPIO ET AL. 11 risk factors. Hence the term ‘metabolically healthy’ may be a misnomer. CONFLICT OF INTEREST No conflict of interest was declared by the authors in relation to the research, authorship and publication of this study. All authors declared that there is no support from or financial relationship with any organization that might have an interest in the submitted work in the previous 36 months and that there is no other rela- tionships or activities that could appear to have influenced the submitted work. AUTHOR CONTRIBUTIONS Mark McEvoy (PhD), A/Professor of Epidemiology: lead investigator, planning and supervising progress of study; one of the independent investigators who conducted quality assessment of the studies; wrote and reviewed the paper from draft stage to completion. Katie Wynne (PhD), A/Professor of Endocrinology: planning and supervising pro- gress of the study; third investigator who was consulted for resolution of any disagreement during data search; reviewed the paper from draft stage to completion. Emma Croker (MBBS), Advanced Trainee in Endocrinology: planning the study, literature search, reviewed litera- ture; one of the independent investigators who analysed citations for relevant studies and conducted data extraction; reviewed the paper from draft stage to completion. John Attia (PhD), Professor of Medi- cine and Epidemiology: reviewed the paper from draft stage to com- pletion. George S Odongo (MPH), Statistician: performed data analysis. Jacob Opio (Master Clinical Epidemiology), PhD Candidate: planning the study including drawing study protocol; made search terms, performed literature search and review; one of the indepen- dent investigators who analysed citations, extracted data and con- ducted quality assessment, wrote and reviewed the paper from draft stage to completion. ORCID Jacob Opio https://orcid.org/0000-0002-2235-9415 REFERENCES 1. Collaboration NRF. Worldwide trends in children's and adolescents' body mass index, underweight, overweight and obesity, in compari- son with adults, from 1975 to 2016: a pooled analysis of 2416 population-based measurement studies with 128.9 million partici- pants. Lancet. 2017;390(10113):2627-2642. 2. Haffner SM. Relationship of metabolic risk factors and development of cardiovascular disease and diabetes. Obesity (Silver Spring). 2006;14 (S6):121S-127S. 3. Stefan N, Häring H-U, Hu FB, Schulze MB. Metabolically healthy obe- sity: epidemiology, mechanisms, and clinical implications. Lancet Dia- betes Endocrinol. 2013;1(2):152-162. 4. Stefan N, Kantartzis K, Machann J, et al. Identification and characteri- zation of metabolically benign obesity in humans. Obstet Gynecol Surv. 2009;64(1):30-31. 5. Wildman RP, Muntner P, Reynolds K, et al. The obese without car- diometabolic risk factor clustering and the normal weight with car- diometabolic risk factor clustering: prevalence and correlates of 2 phenotypes among the US population (NHANES 1999-2004). Arch Intern Med. 2008;168(15):1617-1624. 6. Appleton SL, Seaborn CJ, Visvanathan R, et al. Diabetes and cardio- vascular disease outcomes in the metabolically healthy obese pheno- type: a cohort study. Diabetes Care. 2013;36(8):2388-2394. 7. Hinnouho GM, Czernichow S, Dugravot A, et al. Metabolically healthy obesity and the risk of cardiovascular disease and type 2 diabetes: the Whitehall II cohort study. Eur Heart J. 2015;36(9):551-559. 8. Phillips CM. Metabolically healthy obesity: definitions, determinants and clinical implications. Rev Endocr Metab Disord. 2013;14(3): 219-227. 9. Fan J, Song Y, Chen Y, Hui R, Zhang W. Combined effect of obesity and cardio-metabolic abnormality on the risk of cardiovascular dis- ease: a meta-analysis of prospective cohort studies. Int J Cardiol. 2013;168(5):4761-4768. 10. Kramer CK, Zinman B, Retnakaran R. Are metabolically healthy over- weight and obesity benign conditions?: a systematic review and meta-analysis. Ann Intern Med. 2013;159(11):758-769. 11. Eckel N, Meidtner K, Kalle-Uhlmann T, Stefan N, Schulze MB. Meta- bolically healthy obesity and cardiovascular events: a systematic review and meta-analysis. Eur J Prev Cardiol. 2015;23(9):956-966. 12. Zheng R, Zhou D, Zhu Y. The long-term prognosis of cardiovascular disease and all-cause mortality for metabolically healthy obesity: a systematic review and meta-analysis. J Epidemiol Community Health. 2016;70(10):1024-1031. 13. Ortega FB, Cadenas-Sanchez C, Migueles JH, et al. Role of physical activity and fitness in the characterization and prognosis of the meta- bolically healthy obesity phenotype: a systematic review and meta- analysis. Prog Cardiovasc Dis. 2018;61(2):190-205. 14. YehT-L, Chen H-H, Tsai S-Y, Lin C-Y, Liu S-J, Chien K-L. The relation- ship between metabolically healthy obesity and the risk of cardiovas- cular disease: a systematic review and meta-analysis. J Clin Med. 2019;8(8):1228. https://doi.org/10.3390/jcm8081228. 15. Katzmarzyk PT, ChurchTS, Janssen I, Ross R, Blair SN. Metabolic syn- drome, obesity, and mortality: impact of cardiorespiratory fitness. Diabetes Care. 2005;28(2):391-397. 16. Bo S, Musso G, Gambino R, et al. Prognostic implications for insulin- sensitive and insulin-resistant normal-weight and obese individuals from a population-based cohort. Am J Clin Nutr. 2012;96(5):962-969. 17. Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. JAMA. 2000;283 (15):2008-2012. 18. Page MJ, Shamseer L, Tricco AC. Registration of systematic reviews in PROSPERO: 30,000 records and counting. Syst Rev. 2018;7(1):1–9. 19. Kim SY, Park JE, Lee YJ, et al. Testing a tool for assessing the risk of bias for nonrandomized studies showed moderate reliability and promising validity. J Clin Epidemiol. 2013;66(4):408-414. 20. Caleyachetty R, Thomas GN, Toulis KA, et al. Metabolically healthy obese and incident cardiovascular disease events among 3.5 million men and women. J Am Coll Cardiol. 2017;70(12):1429-1437. 21. Lassale C, Tzoulaki I, Moons KGM, et al. Separate and combined asso- ciations of obesity and metabolic health with coronary heart disease: a pan-European case-cohort analysis. Eur Heart J. 2017;39(5): 397-406. 22. Dhana K, Koolhaas CM, van Rossum EF, et al. Metabolically healthy obesity and the risk of cardiovascular disease in the elderly popula- tion. PLoS ONE. 2016;11(4):e0154273. 23. Hosseinpanah F, Barzin M, Sheikholeslami F, Azizi F. Effect of differ- ent obesity phenotypes on cardiovascular events in Tehran Lipid and Glucose Study (TLGS). Am J Cardiol. 2011;107(3):412-416. 24. Aung K, Lorenzo C, Hinojosa MA, Haffner SM. Risk of developing dia- betes and cardiovascular disease in metabolically unhealthy normal- weight and metabolically healthy obese individuals. J Clin Endocrinol Metab. 2014;99(2):462-468. 12 OPIO ET AL. https://orcid.org/0000-0002-2235-9415 https://orcid.org/0000-0002-2235-9415 https://doi.org/10.3390/jcm8081228 25. Hansen L, Netterstrom MK, Johansen NB, et al. Metabolically healthy obesity and ischemic heart disease: a 10-year follow-up of the inter99 study. J Clin Endocrinol Metabol. 2017;102(6):1934-1942. 26. Lee H-J, Choi E-K, Lee S-H, Kim Y-J, Han K-D, Oh S. Risk of ischemic stroke in metabolically healthy obesity: a nationwide population- based study. PLoS ONE. 2018;13(3):e0195210. 27. Lu J, Bi Y, Wang T, et al. The relationship between insulin-sensitive obesity and cardiovascular diseases in a Chinese population: results of the REACTION study. Int J Cardiol. 2014;172(2):388-394. 28. Meigs JB, Wilson PW, Fox CS, et al. Body mass index, metabolic syn- drome, and risk of type 2 diabetes or cardiovascular disease. J Clin Endocrinol Metab. 2006;91(8):2906-2912. 29. Mirzaei B, Abdi H, Serahati S, et al. Cardiovascular risk in different obesity phenotypes over a decade follow-up: Tehran Lipid and Glu- cose Study. Atherosclerosis. 2017;258:65-71. 30. Morkedal B, Vatten LJ, Romundstad PR, Laugsand LE, Janszky I. Risk of myocardial infarction and heart failure among metabolically healthy but obese individuals: HUNT (Nord-Trondelag Health Study), Nor- way. J Am Coll Cardiol. 2014;63(11):1071-1078. 31. Ortega FB, Lee DC, Katzmarzyk PT, et al. The intriguing metabolically healthy but obese phenotype: cardiovascular prognosis and role of fitness. Eur Heart J. 2013;34(5):389-397. 32. Thomsen M, Nordestgaard BG. Myocardial infarction and ischemic heart disease in overweight and obesity with and without metabolic syndrome. JAMA Intern Med. 2014;174(1):15-22. 33. van der A DL, Nooyens AC, van Duijnhoven FJ, Verschuren MM, Boer JM. All-cause mortality risk of metabolically healthy abdominal obese individuals: the EPIC-MORGEN study. Obesity (Silver Spring). 2014;22(2):557-564. 34. Loprinzi PD, Frith E. Cardiometabolic healthy obesity paradigm and all-cause mortality risk. Eur J Intern Med. 2017;43:42-45. 35. Eckel N, Li Y, Kuxhaus O, Stefan N, Hu FB, Schulze MB. Transition from metabolic healthy to unhealthy phenotypes and association with cardiovascular disease risk across BMI categories in 90 257 women (the Nurses' Health Study): 30 year follow-up from a prospective cohort study. Lancet Diabetes Endocrinol. 2018;6(9):714-724. 36. Song Y, Manson JE, Meigs JB, Ridker PM, Buring JE, Liu S. Compari- son of usefulness of body mass index versus metabolic risk factors in predicting 10-year risk of cardiovascular events in women. Am J Cardiol. 2007;100(11):1654-1658. 37. Twig G, Gerstein HC, Shor DBA, et al. Coronary artery disease risk among obese metabolically healthy young men. Eur J Endocrinol. 2015;173(3):305-312. 38. Arnlov J, Ingelsson E, Sundstrom J, Lind L. Impact of body mass index and the metabolic syndrome on the risk of cardiovascular disease and death in middle-aged men. Circulation. 2010;121(2):230-236. 39. Expert Panel on Detection E. Executive summary of the third report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). JAMA. 2001;285(19):2486–2497. 40. Alberti KGMM, Zimmet P, Shaw J. Metabolic syndrome—a new world-wide definition. A consensus statement from the international diabetes federation. Diabet Med. 2006;23(5):469-480. 41. Ruland S, Hung E, Richardson D, Misra S, Gorelick PB, African Ameri- can Antiplatelet Stroke Prevention Study Investigators. Impact of obesity and the metabolic syndrome on risk factors in African Ameri- can stroke survivors: a report from the AAASPS. Arch Neurol. 2005;62 (3):386-390. 42. Lavie CJ, Laddu D, Arena R, Ortega FB, Alpert MA, Kushner RF. Reprint of: healthy weight and obesity prevention: JACC health pro- motion series. J Am Coll Cardiol. 2018;72(23):3027-3052. 43. Moholdt T, Lavie CJ, Nauman J. Interaction of physical activity and body mass index on mortality in coronary heart disease: data from the Nord-Trøndelag Health Study. Am J Med. 2017;130(8):949-957. 44. Moholdt T, Lavie CJ, Nauman J. Sustained physical activity, not weight loss, associated with improved survival in coronary heart dis- ease. J Am Coll Cardiol. 2018;71(10):1094-1101. SUPPORTING INFORMATION Additional supporting information may be found online in the Supporting Information section at the end of this article. How to cite this article: Opio J, Croker E, Odongo GS, Attia J, Wynne K, McEvoy M. Metabolically healthy overweight/ obesity are associated with increased risk of cardiovascular disease in adults, even in the absence of metabolic risk factors: A systematic review and meta-analysis of prospective cohort studies. Obesity Reviews. 2020;1–13. https://doi.org/10.1111/ obr.13127 OPIO ET AL. 13 https://doi.org/10.1111/obr.13127 https://doi.org/10.1111/obr.13127 Metabolically healthy overweight/obesity are associated with increased risk of cardiovascular disease in adults, even in th... INTRODUCTION METHODS AND MATERIAL Data source and searches Study selection Quality assessment Data extraction Data analysis RESULTS Study characteristics MHOW and risk of cardiovascular events MHO and risk of CVD Investigating heterogeneity and publication bias DISCUSSION RECOMMENDATIONS CONCLUSIONS CONFLICT OF INTEREST AUTHOR CONTRIBUTIONS REFERENCES << /ASCII85EncodePages false /AllowTransparency false /AutoPositionEPSFiles false /AutoRotatePages /None /Binding /Left /CalGrayProfile (Dot Gain 20%) /CalRGBProfile (sRGB IEC61966-2.1) /CalCMYKProfile (U.S. Web Coated \050SWOP\051 v2) /sRGBProfile (sRGB IEC61966-2.1) /CannotEmbedFontPolicy /Error /CompatibilityLevel 1.3 /CompressObjects /Off /CompressPages true /ConvertImagesToIndexed true /PassThroughJPEGImages true /CreateJobTicket false /DefaultRenderingIntent /Default /DetectBlends false /DetectCurves 0.1000 /ColorConversionStrategy /LeaveColorUnchanged /DoThumbnails false /EmbedAllFonts true /EmbedOpenType false /ParseICCProfilesInComments true /EmbedJobOptions true /DSCReportingLevel 0 /EmitDSCWarnings false /EndPage -1 /ImageMemory 1048576 /LockDistillerParams true /MaxSubsetPct 100 /Optimize false /OPM 1 /ParseDSCComments true /ParseDSCCommentsForDocInfo true /PreserveCopyPage false /PreserveDICMYKValues true /PreserveEPSInfo false /PreserveFlatness true /PreserveHalftoneInfo false /PreserveOPIComments false /PreserveOverprintSettings true /StartPage 1 /SubsetFonts true /TransferFunctionInfo /Apply /UCRandBGInfo /Remove /UsePrologue false /ColorSettingsFile () /AlwaysEmbed [ true ] /NeverEmbed [ true ] /AntiAliasColorImages false /CropColorImages true /ColorImageMinResolution 300 /ColorImageMinResolutionPolicy /OK /DownsampleColorImages false /ColorImageDownsampleType /Bicubic /ColorImageResolution 300 /ColorImageDepth 8 /ColorImageMinDownsampleDepth 1 /ColorImageDownsampleThreshold 1.50000 /EncodeColorImages true /ColorImageFilter /FlateEncode /AutoFilterColorImages false /ColorImageAutoFilterStrategy /JPEG /ColorACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /ColorImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000ColorACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000ColorImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 300 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages false /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth 8 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /FlateEncode /AutoFilterGrayImages false /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /GrayImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000GrayACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000GrayImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages false /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict << /K -1 >> /AllowPSXObjects false /CheckCompliance [ /PDFX1a:2001 ] /PDFX1aCheck true /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError false /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (Euroscale Coated v2) /PDFXOutputConditionIdentifier (FOGRA1) /PDFXOutputCondition () /PDFXRegistryName (http://www.color.org) /PDFXTrapped /False /CreateJDFFile false /Description << /CHS /CHT /DAN /DEU /ESP /FRA /ITA (Utilizzare queste impostazioni per creare documenti Adobe PDF che devono essere conformi o verificati in base a PDF/X-1a:2001, uno standard ISO per lo scambio di contenuto grafico. Per ulteriori informazioni sulla creazione di documenti PDF compatibili con PDF/X-1a, consultare la Guida dell'utente di Acrobat. I documenti PDF creati possono essere aperti con Acrobat e Adobe Reader 4.0 e versioni successive.) /JPN /KOR /NLD (Gebruik deze instellingen om Adobe PDF-documenten te maken die moeten worden gecontroleerd of moeten voldoen aan PDF/X-1a:2001, een ISO-standaard voor het uitwisselen van grafische gegevens. Raadpleeg de gebruikershandleiding van Acrobat voor meer informatie over het maken van PDF-documenten die compatibel zijn met PDF/X-1a. De gemaakte PDF-documenten kunnen worden geopend met Acrobat en Adobe Reader 4.0 en hoger.) /NOR /PTB /SUO /SVE /ENG (Modified PDFX1a settings for Blackwell publications) /ENU (Use these settings to create Adobe PDF documents that are to be checked or must conform to PDF/X-1a:2001, an ISO standard for graphic content exchange. For more information on creating PDF/X-1a compliant PDF documents, please refer to the Acrobat User Guide. Created PDF documents can be opened with Acrobat and Adobe Reader 4.0 and later.) >> /Namespace [ (Adobe) (Common) (1.0) ] /OtherNamespaces [ << /AsReaderSpreads false /CropImagesToFrames true /ErrorControl /WarnAndContinue /FlattenerIgnoreSpreadOverrides false /IncludeGuidesGrids false /IncludeNonPrinting false /IncludeSlug false /Namespace [ (Adobe) (InDesign) (4.0) ] /OmitPlacedBitmaps false /OmitPlacedEPS false /OmitPlacedPDF false /SimulateOverprint /Legacy >> << /AddBleedMarks false /AddColorBars false /AddCropMarks false /AddPageInfo false /AddRegMarks false /ConvertColors /ConvertToCMYK /DestinationProfileName () /DestinationProfileSelector /DocumentCMYK /Downsample16BitImages true /FlattenerPreset << /PresetSelector /HighResolution >> /FormElements false /GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks false /IncludeInteractive false /IncludeLayers false /IncludeProfiles false /MultimediaHandling /UseObjectSettings /Namespace [ (Adobe) (CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /DocumentCMYK /PreserveEditing true /UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling /UseDocumentProfile /UseDocumentBleed false >> ] >> setdistillerparams << /HWResolution [2400 2400] /PageSize [612.000 792.000] >> setpagedevice