Statistics from Altmetric.com
In addition to its own potential role as an independent coronary heart disease (CHD) risk factor, substantial evidence indicates that overweightness and obesity adversely impact many other CHD risk factors, including dyslipidaemia, hypertension/left ventricular hypertrophy, glucose intolerance, including metabolic syndrome and type 2 diabetes mellitus and physical inactivity.1 Clearly, the prevalence of almost every cardiovascular disease (CVD), including hypertension, heart failure (HF), peripheral arterial disease, atrial fibrillation as well as CHD, is markedly increased in the presence of obesity. However, despite these adverse effects of obesity, numerous studies and even large meta-analyses clearly demonstrate an ‘obesity paradox’, in which lean patients with established CVD have a worse clinical prognosis than do their more overweight/obese counterparts with the same CVD, including CHD.1 ,2
Despite the obesity paradox, several studies suggest that within the obese cohort, ‘heavier’ obese patients do not have a more favourable prognosis. In fact, in a large systematic review of over 250 000 CHD patients in 40 cohort studies followed for 3.8 years, Romero-Corral and colleagues2 reported that overweight and obese CHD patients had a lower cardiovascular and total mortality compared with underweight and normal-weight CHD patients; however, patients with class II obesity (body mass index (BMI) >35 kg/m2) were at excess risk of cardiovascular mortality but still no increase in total mortality. A recent report by Das et al3 showed a U-shaped inhospital mortality curve in a very large cohort with ST-segment myocardial infarction. Interestingly, those authors classified the class I obese (BMI 30–35 kg/m2) as the reference group and even suggested that in our ‘ever growing’ society, mild obesity may now be the ‘new normal’. Most importantly, however, is the demonstration that although the normal-BMI group had the highest unadjusted inhospital mortality, the patients with class III obesity (BMI ≥40 kg/m2) had increased mortality in both the unadjusted and adjusted models compared with the ‘reference group’ of class I obesity. In a very recent study of over 60 000 patients referred for coronary angiography, compared to a lean reference group (BMI 21.0–23.5 kg/m2), the highest mortality occurred in the underweight (BMI <18.5 kg/m2), the lowest mortality occurred in the modest overweight group (BMI 26.5–28 kg/m2), whereas mortality remained decreased with obesity up to 35 kg/m2, and then increased.4
In the present report, Azimi and colleagues5 extended these findings in a very large cohort of 37 573 patients with documented CHD who were followed for 11 years (median 3.2 years). As reported in many other studies, they demonstrated that the highest mortality occurred in the underweight (BMI <18.5 kg/m2), followed by their large cohort (N=844) of class III obese (BMI ≥40 kg/m2). They also found an increased mortality risk among the low normal BMI group (BMI 18.5–23 kg/m2), whereas the lowest mortality was in the pre-obese (BMI 27.5–30 kg/m2), followed by the mild overweight (BMI 25–27.5 kg/m2). These authors, therefore, suggested that this relationship between BMI and mortality in CHD may represent an ‘overweight paradox’ rather than an ‘obesity paradox’. In fact, even in the general population, a very recent systematic review and meta-analysis of 97 studies of more than 2.88 million individuals demonstrated that the lowest mortality occurred in overweight (BMI 25–30 kg/m2), followed by mild or class I obese (BMI 30–35 kg/m2), supporting the theory of the potential protective association of overweightness in modern times, in both the general population as well as in cohorts with CHD.6
One could argue that a major limitation of most of those studies is the use of BMI as a surrogate measure of body fatness, because BMI increases with both body fat (BF) and fat-free mass (FFM) and thus is only an aggregate measure of body composition. Direct measures of body composition potentially include waist circumference to assess central obesity and direct BF and FFM measurements, all or some of which may provide a superior risk assessment of obesity than BMI. It stands to reason that increasing BF may adversely impact health, whereas increases in FFM could be beneficial. Nevertheless, almost a decade ago, we demonstrated the obesity paradox in HF using both high BMI and BF, both of which predicted better survival.7 In fact, we demonstrated that for every 1% increase in BF, survival in advanced HF increased by 13%. Years later, we confirmed that CHD patients with high BF also had better subsequent survival.8
We recently suggested that this paradox may actually represent a ‘lean paradox’ as opposed to an ‘obesity paradox’.9 In a study of 581 CHD patients followed on average for 3 years, the group with both low BMI (<25 kg/m2) and low BF (<25% in men and <35% in women) had highest mortality (11% over 3 years), an observation that remained even after excluding the underweight (BMI <18.5 kg/m2). The other groups divided by high/low BMI and BF had lower mortality risks (figure 1).9 In this cohort, low BF predicted a 3.5-fold higher 3-year mortality, supporting the concept of a ‘lean paradox’. Most recently, we extended this relationship assessing both BF and FFM (or lean mass). We demonstrated that the highest mortality occurred in patients with both low BF and low FFM and the lowest mortality occurred in those with both high BF and high FFM, whereas the other two groups had intermediate mortality risks (figure 2).10 In fact, a low BF increased mortality risk by 2.6-fold whereas a low FFM or lean mass increased mortality by 3.1-fold. These data support the concept that body composition besides BMI predicts mortality in CHD, as well as the fact that the leanest patients, even excluding the underweight, with both low BF and FFM, had the worst prognosis.
A second factor not assessed in most studies of the obesity paradox in CVD and CHD has been the impact of cardiorespiratory fitness (CRF). Certainly, CRF is an often forgotten cardiovascular risk factor that significantly mitigates the obesity paradox.11 In fact, the common theme from three recent papers was that CHD patients with high CRF not only had a better prognosis, but also did not have an obesity paradox, regardless of how they defined obesity (BMI,12–14 BF,14 and even waist circumference or central obesity14). The paradox was only present in those with low CRF. This relationship between BMI, CRF and mortality has also recently been described in patients with HF.15 In all of these studies, the worst prognosis was in the lean patients with low levels of CRF, demonstrating that CRF significantly modifies the relationship between BMI and survival in both CHD and HF, and thus assessing the obesity, overweight or lean paradox, however it is referred to, may be misleading unless CRF is also considered.14 ,15
Finally, in light of this paradox (obesity, lean, or overweight paradox as suggested by Azimi and colleagues5) as well as the lack of hard evidence, especially from large randomised trials, of weight loss on protecting against major clinical events, some have questioned the value of weight loss in patients with CVD, including CHD.16 Although we still believe that the ‘weight’ of evidence supports purposeful weight loss, as is supported in many mostly small studies of CHD patients,1 ,8 ,17–19 efforts to improve FFM and physical fitness, including both CRF and muscular strength11 ,20 could one day prove better advice in select groups than weight loss. The best way to do this is with exercise training, both aerobic and resistance training, especially within the often underutilised cardiac rehabilitation and exercise training programmes for CHD.11 ,20 Certainly, greater efforts are needed to promote physical fitness and phase II cardiac rehabilitation and exercise training programmes in secondary CHD prevention.
Competing interests None.
Provenance and peer review Commissioned; internally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.