Abstract
Objectives
There is an established causal link between obesity and cardiovascular outcomes. The aim of this review was to determine whether an independent relationship exists between anthropometric measurements of weight (typically body mass index [BMI]) and cardiovascular outcomes (e.g. angina, myocardial infarction, congestive heart failure, stroke, and mortality due to cardiovascular disease) in the general population and in patients with type 2 diabetes.
Methods
A review of the medical literature published between 1988 and May 2008 was conducted using the PubMed, EMBASE, Cochrane and Center for Review and Dissemination databases. Studies longer than 12 months, with ≥500 adult subjects and published in English were included.
Results
In studies conducted in general populations there was an overall trend towards increased risk for adverse cardiovascular outcomes with increasing BMI. The nature and strength of this relationship varied according to the measurement used (e.g. BMI, waist circumference, waist-to-hip ratio) and the population studied, with notable differences observed in Asian/Asia-Pacific compared with European or North American-based studies. However, data from diabetes-specific populations are limited.
Conclusions
In general, the degree of being overweight or obese was associated with an elevated risk of adverse cardiovascular events and mortality. Although inextricable links exist between obesity, type 2 diabetes and cardiovascular disease in the general population, the extent to which findings can be extrapolated to a diabetes-specific population is limited.
Introduction
Type 2 diabetes and obesity both represent significant, and growing, global public health problems. Recent estimates suggest that in the region of 900 million to 1.6 billion adults worldwide are classified as overweight (body mass index [BMI] ≥ 25 kg/m2) and that approximately 400 million adults are obese (BMI ≥ 30 kg/m2).1,2 Moreover, by the year 2015 the World Health Organization (WHO) project that these figures will rise to approximately 2.3 billion overweight and 700 million obese adults 1 and by 2030 (assuming current secular trends continue) Kelly et al estimate the global prevalence of overweight and obesity will be 2.16 billion and 1.12 billion, respectively. 2 The estimated prevalence of diabetes is equally alarming; Wild et al 3 estimated a global prevalence of 171 million in 2000 rising to 366 million in 2030, with the number of diabetes-related deaths projected to increase by more than 50% over the next 10 years. 1 Notably, the projections of Wild et al, 3 assume no increase in the prevalence of obesity and may therefore represent a conservative scenario.
Type 2 diabetes and obesity were once generally considered to be health problems that existed almost exclusively within high-income Westernized countries. However, the burden of diabetes and obesity is increasing at an alarming rate in many low- and middle-income countries and it is estimated that by 2025 three quarters of all people with diabetes will reside in developing countries. 4 Moreover, type 2 diabetes and obesity share a number of common etiological factors and a large part of the ever increasing burden of both conditions has been attributed to a shift towards an increased intake of energy dense foods with a high fat and sugar content combined with increasingly sedentary lifestyles. 1 A number of studies have shown that obesity is an independent risk factor for the development of type 2 diabetes5–8 and in the US an estimated 60%–80% of diabetes patients are classified as abdominally obese. 9 Additionally, type 2 diabetes is an established risk factor for cardiovascular disease and it is estimated that individuals with diabetes are 2–6 times more likely to die from cardiovascular causes compared with those without diabetes. 10 Given the intricate nature of the association between obesity and cardiovascular disease and type 2 diabetes it is generally assumed that obesity should be associated with an elevated risk of cardiovascular disease in patients with type 2 diabetes. However, there is a general paucity of data from long-term longitudinal studies that examine the link between anthropometric measurements of overweight/obesity and cardiovascular events specifically within type 2 diabetes populations. We performed a literature review with the primary objective of elucidating whether BMI or other anthropometric measurements of body weight and adiposity are independent risk factors for cardiovascular events in patients with type 2 diabetes. However, owing to the lack of data from diabetes-specific populations the review also examined studies conducted in general populations, where studies were examined to ascertain whether diabetes was adjusted for as a covariate.
Methods
A literature search was performed in May 2008 to identify English language articles published within the last 20 years relating to the association between obesity and cardiovascular outcomes. Specifically, studies reporting results in terms of clinically relevant endpoints rather than association between risk factors were sought. Searches were performed using PubMed, MEDLINE, EMBASE, the Cochrane database, and the Center for Review and Dissemination databases. We included observational, prospective and retrospective longitudinal studies performed in adults (both single sex and mixed sex studies as well as studies performed in populations with pre-existing cardiovascular conditions were included); pooled- and meta-analyses were also included. Additionally, studies were required to have in excess of 500 patients and if reference to a follow-up period was made in the article this was required to be longer than 12 months. In addition to the online literature searches, reference sections of retrieved articles were examined in order to identify additional relevant articles. Article titles and abstracts were initially reviewed followed by full-text review where necessary. Owing to the overlapping nature of cardiovascular endpoints used, for example, coronary heart disease (CHD) and myocardial infarction (MI); where CHD was often specifically defined as non-fatal MI or death from CHD, and also to avoid potential misinterpretation of findings, studies were classified and are discussed according to endpoints as presented in the articles.
Results
The literature review identified a total of 51 studies that were examined in detail (Table 1). Endpoints assessed in these studies included incidence of MI (n = 9), angina (n = 1) ischemic heart disease (IHD; n = 2), incidence of and mortality from CHD (n = 11), heart failure and outcomes in patients with existing heart failure (n = 4) and overall incidence/prevalence of cardiovascular disease and mortality due to cardiovascular disease (n = 20).
Summary findings of studies examining the relationship between anthropometric indices of obesity and cardiovascular endpoints.
Denotes studies either conducted in diabetes patients or adjusting for diabetes.
Myocardial infarction and ischemic heart disease
Overall, a total of eight studies reported here examined the relationship between overweight/obesity and the incidence of MI, IHD or angina11–17 (one study investigated the link between obesity and age at incidence of MI). 18 In addition, five studies investigated the relationship between obesity and endpoints including recurrent MI, overall mortality and CHD mortality in patients with previous MI.19–23 All studies reported here that examined the link between overweight/ obesity and MI, angina or IHD used BMI as an index of obesity; however, a substantial proportion of studies also used other indices of obesity including waist-to-hip ratio (WHR), waist circumference (WC) and visceral adipose tissue area.
Overall, there was a general trend towards a positive association between increased BMI and elevated risk for MI, although in some studies the strength of the association was such that it was not statistically significant. In a study of over 20,000 men enrolled in the Physician's Health Study Rexrode et al 11 reported a significant association between increasing BMI, by quintile, and the risk of MI (Table 1). Men in the highest BMI quintile (≥27.6 kg/m2) had a relative risk (RR) for MI of 2.52 compared with men with a BMI < 22.8 kg/m2. The association between BMI and risk for MI was apparent (and statistically significant) following adjustment for risk factors including smoking, history of MI, alcohol consumption, physical activity and WHR quintile. However, these data were not adjusted for diabetes and Rexrode et al 24 note that the men enrolled in the Physician's Health Study were generally healthier and had lower rates of obesity compared with the general population, which in turn, limits the interpretation of these results in a general population.
A graded association between increased BMI and the risk of MI was also observed in a large case-control study conducted by Yusuf et al 13 although this association became non-significant following adjustment for other risk factors including activity, alcohol use, diet, smoking, history of hypertension and history of diabetes. However, the authors reported a significant association between increased WHR and elevated risk for MI that remained following adjustment for other risk factors. Further, not only do these findings suggest that WHR may be a better predictor for MI risk compared with BMI, Yusuf et al also postulate that the reliance on BMI as a anthropometric measure may lead to an underestimation of the global burden of obesity, particularly in the Middle East and South and Southeast Asia. 13
In a study of 2,503 elderly men and women Nicklas et al 12 observed a non-significant relationship between a 4.90 kg/m2 increase in BMI and risk of MI, although in women a significant relationship was evident between visceral adipose tissue area and the risk of MI. Interestingly, using data collected from the Framingham Heart Study Wilson et al 17 reported that overweight women had a lower risk of MI compared with normal weight women (BMI 18.5–24.9 kg/m2) but obese women had an elevated relative risk RR (95% confidence interval [CI]) of 1.46 (0.94–2.28) compared with normal weight women (following adjustment for risk factors including diabetes, smoking, hypertension and hypercholesterolemia). In men, Wilson et al 17 reported a lower risk of MI for obese men compared with overweight men (although both categories had an elevated risk for MI compared with men with a BMI in the normal range). In the same study the authors also investigated the risk for angina and reported an increased risk for angina with increasing BMI (following multivariate adjustment, including for diabetes) in both men and women. In common with many investigators, Wilson et al excluded underweight (BMI < 18.5 kg/m2) subjects from their analysis–-stating that this group typically contains a high proportion of heavy smokers, and subjects with severe chronic disease or malignancies.
Newton and LaCroix 19 examined the rate of reinfarction following first MI in women and reported a 5% increase in the risk of reinfarction per unit increase in BMI, although this association was attenuated, but not eliminated, following adjustment for diabetes and hypertension. Rea et al 22 investigated the risk for recurrent coronary events following first MI. They found that moderately overweight (BMI 25.0–27.4 kg/m2) subjects had a lower relative risk for recurrent coronary events compared with subjects with a normal BMI (16–24.9 kg/m2) but risk increased in a stepwise fashion in severely overweight subjects (27.5–29.9 kg/m2) and those with class I and class II obesity (30–34.9 kg/m2 and ≥35 kg/m2, respectively). 22 Further, the authors postulate that approximately 43% of this excess risk could be attributed to clinical measurements of diabetes, hypertension and dyslipidemia. 22 Additionally, Kragelund et al 21 observed a non-significant relationship between obesity and mortality in patients with previous acute MI, although overweight and obese men and women had numerically lower RRs and underweight men and women had a numerically higher RR for mortality compared with their normal weight counterparts. In addition, Ness et al 23 reported a reverse J-shaped association between BMI and CHD mortality in men with a recent history of MI.
Three studies included in this review examined the link between BMI and IHD. Jee et al, 16 Ni Mhurchu et al 15 and the Asia Pacific Cohort Studies Collaboration 14 all reported an increased risk for IHD with increased BMI. Notably, all three of these studies/pooled analyses were performed in Asian or Asian-Pacific populations. Jee et al 16 used a reference group with a BMI of 18–< 19 kg/m2 and reported an elevated risk for IHD in subjects with a BMI above 19 kg/m2. These results are reflective of the fact that overweight and obesity are defined by lower BMI thresholds in Asian populations compared with European and North American populations.
Heart failure
The relationship between obesity and heart failure appears to be complex; obesity is reported to be an independent risk factor for the development of heart failure, 25 yet conversely, in patients with established heart failure, obesity appears to be associated with a survival advantage. Although a number of hypotheses exist, the mechanism underlying this “obesity paradox” has yet to be fully elucidated. Kenchaiah et al 25 investigated the relationship between BMI and heart failure in participants of the Framingham Heart Study (both as a continuous and categorical variable; Table 1) over a follow-up period of up to 14 years. They observed a graded association between increased BMI and the risk of developing heart failure in both men and women (following adjustment for age, smoking status, alcohol intake, total serum cholesterol, valve disease, hypertension, left ventricular hypertrophy, MI and diabetes). Notably, the authors suggest that elevated BMI may predispose certain individuals to heart failure owing to the causal link between elevated BMI and atherogenic traits such as hypertension and type 2 diabetes, which are themselves risk factors for the development of heart failure. Additionally, they discuss the reported links between increased BMI and left-ventricular remodeling, which may be attributable to a number of factors including increased hemodynamic load, neurohormonal activation and oxidative stress.
Bibbins-Domingo et al 26 examined potential risk factors for the development of heart failure in postmenopausal women with established coronary disease; diabetes was found to be an independent risk factor for heart failure in this population but overweight (BMI 25–30 kg/m2) or class I obesity (BMI 30–35 kg/m2) were not. Only a BMI of >35 kg/m2 was found to be an independent risk factor for the development of heart failure in this group of women. 26 However, Bender et al 27 examined the risk for mortality due to heart failure or cardiac dysrhythmias in obese patients and reported a massively elevated risk for mortality in men and women with a BMI ≥ 25 kg/m2 (standardized mortality ratio [95% confidence interval] was 15.57 [10.58–22.10] for men and 8.81 [6.45–11.75] for women compared with the population of Germany).
As stated above, obesity may be a risk factor for the development of heart failure but conversely it appears to be indicative of a more favorable prognosis in terms of overall survival in patients with established heart failure. For example, an inverse correlation between increased BMI and overall mortality was observed by Lissin et al in a group of male US veterans with established heart failure. 28 This “obesity paradox” has also been observed in terms of cardiovascular mortality, for example, Curtis et al 29 observed higher RRs for death from cardiovascular causes and death from worsening heart failure in patients with a normal BMI (18.5–24.9 kg/m2) compared with both underweight (BMI < 18.5 kg/m2), overweight (BMI 25.0–29.9 kg/m2), and obese (BMI ≥ 30 kg/m2) patients. Although diabetes is an established independent risk factor for heart failure and a number of studies performed multivariate adjustment that included diabetes there was a notable paucity of data investigating whether the obesity paradox also exists in overweight or obese patients with concomitant type 2 diabetes and heart failure. Nevertheless, potential explanations for the obesity paradox observed in patients with heart failure include the hypothesis that a proportion of overweight or obese patients who present with cardinal signs and symptoms of heart failure such as dyspnea, leg and ankle edema and basilar pulmonary crepitations may be diagnosed with heart failure without actually having heart failure.25,30 Alternate hypotheses to explain this paradox include the postulate that the increased risk for mortality in patients with a lower BMI may be due to the deleterious effects of cardiac cachexia and muscle wasting associated with advanced heart failure rather than a beneficial effect of increased BMI per se. 29 However, in the study by Curtis et al 29 underweight subjects (BMI < 18.5 kg/m2) also had a lower hazard ratio (HR; 95% CI) for death due to worsening heart failure 0.86 (0.56–1.32) compared with subjects in the normal BMI range (18.5–24.9 kg/m2).
Coronary heart disease
In terms of the relationship between obesity and CHD, an overall trend towards an increased risk of CHD incidence (or mortality due to CHD) with increasing BMI was evident (Table 1).31–36 However, a systematic review performed by Whitlock et al 36 and the results of Canoy et al 37 provide some evidence of a J-shaped relationship between BMI and CHD with Canoy and colleagues reporting that individuals in the lowest BMI quintile had an elevated risk for CHD compared with those in the second-lowest quintile. Interestingly, the same authors also investigated the relationship between WHR and WC and risk for CHD and concluded that indices of abdominal adiposity more predictive of the risk of CHD compared with BMI.
Overweight and obesity are commonly defined as a BMI of 25–29.9 kg/m2 and >30 kg/m2, respectively. However, these classifications are not appropriate for all populations, as illustrated by the findings of Zhang et al 38 who investigated the relationship between BMI and CHD (including non-fatal MI and CHD death) in 67,334 Chinese women. It was reported here that women with a BMI over 22.2 kg/m2 had a more than twofold increase in the RR for CHD compared with those with a BMI below 22.2 kg/m2. These findings suggest that the detrimental health consequences of obesity, in terms of CHD risk, generally commence at a lower BMI threshold in this population compared with predominantly North American or European populations. Furthermore, the findings of Zhang et al 38 are concordant with the findings of studies that investigated the relationship between BMI and IHD in Asian populations.14–16 The differences in the relationship between increased BMI and negative health consequences that exist between Asian populations and North American and European populations are such that the WHO now define overweight as a BMI of 23–25 kg/m2 and obesity as a BMI of ≥25 kg/m2 in Asian populations. 39
Only one study identified during the course of the literature review reported here explicitly examined the relationship between obesity and CHD mortality in a subgroup of patients with diabetes (CHD defined as International Classification of Diseases 9 [ICD-9] codes 410–414). Mann et al 40 reported that for obese non-diabetics the RR (95% CI) for CHD mortality was 1.44 (1.12–1.84) compared with non-diabetic individuals in the normal BMI range (< 25 kg/m2). Unexpectedly, Mann et al observed that overweight non-diabetics had a lower relative risk for CHD mortality compared with non-diabetics in the normal BMI range. For subjects with diabetes the RRs for CHD mortality for each BMI category (lean, overweight and obesity) were more than twofold greater than the RR for the corresponding BMI category for non-diabetics (following adjustment for age, race, smoking, hypertension, serum cholesterol, education level, alcohol consumption and physical activity). In subjects with diabetes there was also a trend for increased risk due to CHD mortality with increasing BMI; indeed in comparison with lean non-diabetics, obese subjects with diabetes had a relative risk (95% CI) of CHD mortality of 3.32 (1.87–5.91). 40
Cardiovascular disease
A large number of studies have investigated the link between increased BMI and clinical endpoints that have been placed under an umbrella term such as cardiovascular disease, circulatory disease or major adverse coronary events (Table 1). A substantial proportion of these studies reported a trend (both significant or non-significant) towards an increased risk in incidence, or mortality due to cardiovascular disease or coronary events with increasing BMI.27,35,41–45 However, as with other clinical endpoints examined here, a J-shaped, U-shaped or inconsistent relationship between BMI and the risk for cardiovascular disease or mortality due to cardiovascular disease was reported in some studies.17,36,46–48 This suggests that patients with low BMI may also have an elevated risk of incidence, or mortality, due to cardiovascular disease compared with individuals with a normal BMI. For example, in a study of over 80,000 US physicians aged 40–84 years Ajani et al 46 reported that men with a BMI < 20 kg/m2 had a higher RR of cardiovascular mortality compared with men with a BMI of 22.5–24.9 kg/m2 (1.50 for BMI < 20 kg/m2 versus 1.00 for BMI of 22.5–24.9 kg/m2; adjusted for smoking, age, alcohol intake and physical activity). They also reported a trend for an increased risk of cardiovascular mortality with increasing BMI in men in BMI categories above 25 kg/m2. A gender-effect was also noted by some investigators, for example Domanski et al 49 reported that in subjects with established coronary artery disease, obesity (BMI ≥ 30 kg/m2 versus BMI < 30 kg/m2) was associated with an increased risk for a composite endpoint of major adverse coronary event in men but not in women. However, in a large prospective study of over 115,000 women enrolled in the Nurses’ Health Study Manson et al 43 reported a trend for an increased age-adjusted RR for mortality due to cardiovascular disease with increasing BMI (in women who were never smokers). Notably, for those women with a BMI over 29 kg/m2 the RR of death from cardiovascular disease was sevenfold greater compared with women with a BMI in the range of 19–21.9 kg/m2. Wilson et al 17 examined the association between BMI and cardiovascular disease and mortality in men and women enrolled in the Framingham Heart Study. It was noted that the risk for total cardiovascular disease increased with increasing BMI in both men and women. However, in terms of the RR of death due to cardiovascular causes overweight women and obese men had a lower risk for death than subjects in the reference category.
In a meta-analysis of 26 studies McGee et al 35 reported a trend for an increased risk of cardiovascular disease with increasing BMI; however, the magnitude of the increased risk was less than that observed by other studies such as the study by Manson et al. 43 McGee et al also reported a RR for cardiovascular mortality of approximately 1.5 in both obese men and women. 35 Moreover, the authors make an interesting point that whilst the relationship between obesity and traditional risk factors for cardiovascular disease such as type 2 diabetes and hypertension is well established the role of non-traditional factors such as homocysteinemia and levels of lipoprotein A are less well characterized.
A number of investigators have considered the effect of overweight/obesity in combination with the metabolic syndrome on clinical outcomes. Katzmarzyk et al 50 compared the risk of mortality due to cardiovascular disease in “healthy” overweight and obese men with that of overweight and obese men with metabolic syndrome. In both healthy men and those with metabolic syndrome the risk of mortality due to cardiovascular disease increased with increasing BMI. However, the increase in risk was higher for overweight and obese men with metabolic syndrome compared with healthy men in the corresponding BMI group. Kip et al 51 also observed an elevated risk for major adverse coronary event with increasing BMI in dysmetabolic women with myocardial ischemia but not in women classified as metabolically normal.
The findings of some investigators suggest that the obesity paradox may not be limited to patients with heart failure. In a retrospective study of 5,950 men and women with known or suspected coronary artery disease Galal et al 52 reported that the hazard ratio (HR) for cardiac death or death from acute MI in underweight patients (BMI < 18.5 kg/m2) was more than double that for subjects in the normal BMI range (18.5–24.9 kg/m2). In comparison, overweight (BMI 25–29.9 kg/m2) and obese (BMI ≥ 30 kg/m2) subjects had a lower HR for cardiac death compared with subjects in the normal BMI range (following multivariate adjustment). 52 In a study conducted in Micronesian Naurans and Melanesian and Indian Fijians Hodge et al 53 reported that there was little evidence to suggest that obesity is a risk factor for cardiovascular disease in these populations, although curiously there was an inverse correlation between overall mortality and obesity in some diabetic sub-groups analyzed. Nauru is known to have among the highest prevalence of diabetes in the world, which makes the observations of Hodge et al all the more pertinent. However, it is possible that BMI may not be the most appropriate measure of obesity in this population. A study by Rush et al 54 compared percentage body fat and BMI in women from New Zealand and Polynesia; they found that at a given percentage body fat Polynesian women typically had a BMI approximately 3–4 units higher than New Zealand women.
Discussion
The initial aim of this review was to examine the link between BMI and adverse cardiovascular endpoints specifically in patients with type 2 diabetes. However, until recently there has been a notable paucity of data from this particular group. A recent study by Eeg-Olofsson et al 55 examined the association between BMI at baseline (as a continuous variable) and first-incident fatal or non-fatal CHD, stroke and cardiovascular disease in 13,087 patients with type 2 diabetes enrolled in the Swedish National Diabetes Register. It was reported that there was a significant relationship between BMI and CHD, stroke and cardiovascular disease (adjusted for age, gender, duration of diabetes, type of hypoglycemic treatment, smoking and significant interaction variables) and we discuss the findings of Eeg-Olofsson et al in the context of this review in detail below.
The underlying trend emerging from studies conducted in general populations and included in this review was, perhaps unsurprisingly, that overweight and obese subjects are at an elevated risk for cardiovascular disease or mortality due to cardiovascular causes compared with subjects with a normal BMI. The results from a number of studies also suggest that underweight patients are at an elevated risk for cardiovascular disease or mortality due to cardiovascular disease compared with their normal weight counterparts. However, when attempting to draw conclusions with regards to overall trends it should firstly be noted that a formal meta-analysis was not performed here and that the heterogeneity of the clinical endpoints, sample populations and methods used complicates the ability to directly compare trends across studies. Additionally, some studies with lengthy follow-up periods excluded the first few years of follow-up in order to reduce confounding from the presence of sub-clinical disease at baseline. In comparison, a number of the studies included here were performed in groups in which established cardiovascular disease (or heart failure or previous MI) were a prerequisite for inclusion.
It is apparent that there are a number of underlying pathophysiological changes that are associated with obesity, which may contribute towards the elevated risk for adverse cardiovascular outcomes. Individuals who are defined as being overweight or obese commonly have a greater proportion of adipose tissue than individuals with a normal BMI. Adipose tissue represents an important endocrine organ and consequently overweight and obese individuals frequently have elevated levels of a number of adipocytokines such as leptin and resistin as well as elevated levels of inflammatory cytokines such as interleukin-6 (IL-6) and tumor necrosis factor alpha (TNF-α). Together, the elevated levels of these cytokines combined with the effects of the dysregulation of downstream targets may contribute towards pathophysiological processes that are frequently manifest in the form of cardiovascular disease as well as other comorbid conditions. Serum leptin levels have been shown to be directly correlated with percentage body fat 56 and increased leptin levels are independently associated with cardiovascular disease and insulin resistance as well as the generation of reactive oxygen species in endothelial cells. 57 The role of resistin, another adipocytokine that is frequently found in elevated levels in obese subjects, is less well characterized but speculation exists that resistin may represent a key component of the mechanistic link between obesity and the development of insulin resistance and type 2 diabetes and cardiovascular disease. 58 In contrast, the secretion of the adipocytokine adiponectin is reduced in both obese individuals and in those with type 2 diabetes compared with healthy age-matched controls. 59 Adiponectin has been linked to MI and cardiovascular risk; in a study of 18,000 men increased circulating levels of adiponectin were linked with a decreased risk of MI. 59 Further, in men with type 2 diabetes increased levels of circulating adiponectin were associated with decreased cardiovascular risk. 59 An inverse correlation has also been observed between circulating levels of adiponectin and a number of cardiovascular risk factors including hypertension, and C-reactive protein (CRP) levels. Adiponectin is also thought to play an important role in a number of steps involved in the development of atherosclerosis as well as vascular and cardiac remodeling processes. 60
There are a number of difficulties in interpreting the data relating anthropometric measurements to cardiovascular risk. The observation that a proportion of the studies included here reported a J-shaped or U-shaped relationship between BMI and adverse cardiovascular endpoints, suggests that underweight patients are also at an elevated risk for cardiovascular disease. As a number of studies either did not include underweight patients or grouped them into the same BMI category as normal weight subjects it is possible that there may be a general underestimation of the detrimental effects of being underweight in terms of cardiovascular disease. Although again, the effects of cachexia due to aging or the presence of comorbid conditions such as chronic obstructive pulmonary disease may be a contributing factor in terms of the elevated risks often observed in underweight patients. These conclusions are then seemingly countered, however, by the “obesity paradox” observed in patients with heart failure and with other preexisting cardiovascular conditions. As discussed earlier, it is possible the deleterious effects of muscle wasting associated with advanced disease or the misdiagnosis of obese patients with signs and symptoms consistent with a differential diagnosis of heart failure may be contributing factors in this paradox but these facts are difficult to reconcile with each other. Interestingly, the obesity paradox appears not to be limited to cardiovascular diseases as increased BMI has also been associated with improved overall survival in patients on maintenance hemodialysis 61 and in patients that have undergone percutaneous coronary intervention. 62
In this review we have investigated the link between BMI and adverse cardiovascular outcomes and although other indices of obesity such as WC and WHR exist, BMI appears to be the most commonly used, possibly due at least in part to its simplicity. However, as an index of obesity BMI is associated with certain obvious caveats that may lead to erroneous conclusions being drawn from studies. Firstly, BMI does not account for body disposition and is therefore prone to overestimating “fatness” in groups such as athletes and underestimating “fatness” in groups in which sarcopenic obesity may be prevalent, such as the elderly. The agreement in terms of the correlation between different indices of obesity and clinical endpoints is also variable and there is a widely held belief that fat distribution rather than absolute amount may be a more important prognostic indicator.12,37,63 In terms of predicting the risk for type 2 diabetes both BMI and measures of central adiposity including WC and WHR are strong independent predictors for the development of the disease.64–66 Additionally, weight gain in adult life is also thought to be an important factor in the risk for developing type 2 diabetes. 64 Another important limitation associated with defining overweight and obesity in terms of BMI is its applicability in Asian populations. In Asian populations the detrimental health effects associated with being overweight or obese commence at a lower BMI compared with European or North American populations. Consequently, in Asian populations a threshold value of 23–25 kg/m2 is commonly applied for the overweight category and ≥25 kg/m2 for obesity, in line a report published by the WHO in 2000. 39 The need for anthropometric measurements and thresholds appropriate to the population under investigation is underscored by the findings of numerous studies. For example, data from a study by Lear et al 67 suggests that at a given BMI percentage body fat is approximately four percent greater in South Asians compared with Europeans. Such observations may potentially have implications for the treatment of types with agents such as the thiazolidinedione, which are associated with weight gain due to expansion of the subcutaneous fat depot (and edema in some patients), whereas the proportion of visceral adipose tissue remains stable (or even decreases). However, the long-term implications of the use of thiazolidinediones in terms of cardiovascular risk remain largely unknown.
The link between increasing BMI and cardiovascular disease has been the subject of extensive investigation, as has the link between increasing BMI and the risk for developing type 2 diabetes. The paucity of published data investigating the relationship between elevated BMI and the risk of cardiovascular disease, MI, heart failure and other endpoints specifically in patients with type 2 diabetes is therefore surprising. This relationship was however the subject of a recent study by Eeg-Olofsson et al 55 where, in an observational study of over 13,000 individuals in the Swedish National Diabetes Register, it was shown that a five unit increase in BMI at baseline was associated with a 15% increase in the risk of CHD, an 11% increase in the risk if stroke and a 13% increase in the risk of cardiovascular disease. Additionally, the authors estimated that the impact of elevated BMI on HbA1c, blood pressure, hyperlipidemia and microalbuminuria was responsible for 40% of the increased risk in terms of CHD and 46% for CVD; they also speculated that the effect of BMI on endothelia dysfunction and low grade inflammation may also be important in the development of cardiovascular disease in overweight and obese patients with diabetes. Moreover, Eeg-Olofsson et al also investigated the impact of weight change during the study on the risk for CHD reporting that a one unit increase in BMI was associated with a 13% increase in the risk for fatal or nonfatal CHD. In terms of the limitations of this study Eeg-Olofsson and colleagues acknowledge that due to lack of data they were unable to adjust for blood lipid values. Additionally, this study was observational in nature and therefore associated with the inherent limitation of this type of study design.
The answers sought for in this review may well be answered by the Look AHEAD (Action for HEAlth in Diabetes) study 68 which is currently ongoing and will examine cardiovascular disease risks and overall mortality in 5,145 obese diabetes patients. Individuals enrolled in this investigation were randomized to either lifestyle intervention or diabetes support and education and will receive 11.5 years of follow up with an anticipated delivery of results by 2012. The Look AHEAD study will also examine the long-term consequences of intentional weight loss and weight gain in this population, a variable that is frequently overlooked in the context of clinical trials and/or epidemiological studies. Although there are data that report the short-term benefits of weight loss in patients with type 2 diabetes, there are little data that examine the long-term implications of weight loss, but the results of the Look AHEAD trial will hopefully address this.
In summary, there is an established causal link between elevated BMI and the increased risk of cardiovascular diseases in general populations, although an obesity paradox in terms of overall survival has been noted in patients with pre-existing cardiovascular conditions and heart failure. Recent evidence from an observational study suggests that overweight and obesity are also independent risk factors for the development of cardiovascular disease in patients with type 2 diabetes.
Grant Supporters
This study was supported by funding from Eli Lilly.
Footnotes
Acknowledgements
This study was funded by Eli Lilly and company.
GG and JSP are current or former employees of IMS Health who have received consultancy fees from Eli Lilly and company; KB and AK are employees and shareholders of Eli Lilly and Company.