editorial
Calcific Aortic Stenosis — Time to Look More Closely at the Valve

Catherine M. Otto, M.D.

Calcific aortic stenosis is a progressive disease that results in stiff valve leaflets with eventual obstruction to left ventricular outflow. Once symptoms occur, valve replacement is the only effective treatment, and there are no known therapies to prevent disease progression. However, several lines of evidence suggest that calcific valve disease is not simply due to age-related degeneration but, rather, is an active disease process with identifiable initiating factors, clinical and genetic risk factors, and cellular and molecular pathways that mediate disease progression.
The key initiating factor in the development of calcific aortic stenosis appears to be mechanical
stress. Specifically, a congenitally bicuspid valve, which is present in about 0.5 to 0.8% of
the population, is the underlying anatomy in the majority of valve replacements for aortic stenosis. [1] Blood-flow dynamics may also play a role, since early lesions are located on the aortic side of the valve in regions with low shear stress.
Clinical factors that are associated with the presence of calcific valve disease include older
age, male sex, elevated serum levels of low-density lipoprotein and lipoprotein(a), smoking, hypertension, diabetes, and the metabolic syndrome.[2]
The presence of mild valve changes, even in the absence of obstruction to blood flow, is associated with an increase of 50% in the risk of myocardial infarction and death from cardiovascular causes during the next 5 years. Genetic factors are difficult to study in a disease that often is not evident until the sixth or seventh decade of life. However, in a subgroup of families, a bicuspid valve appears to be inherited in an autosomal dominant pattern. In one study in France, familial clustering of calcific disease in trileaflet valves also was shown. Mutations in the signaling and transcriptional regulator NOTCH1 gene have been
identified in families with bicuspid aortic valves and leaflet calcification.[3] Case–control studies
have suggested an association between calcific aortic stenosis and genetic polymorphisms in
the vitamin D receptor, estrogen receptor, apolipoprotein E4, and interleukin-10 alleles.
Our understanding of disease progression at the tissue level is based on human valve studies
of either early lesions or end-stage disease, with the assumption that these processes represent
the ends of a disease spectrum .[4,5] Experimental models support this assumption, with
the demonstration that valve lesions occur in the presence of hypercholesterolemia, resulting in
leaflet calcification and valve obstruction.[6] Taken together, the association of calcific aortic stenosis with elevated serum lipid levels, the presence of lipid accumulation in the leaflets, and the increased risk of atherosclerotic clinical end points all lead to the hypothesis that lipid lowering therapy might slow or prevent disease progression. This hypothesis was supported by several retrospective clinical studies indicating slower hemodynamic progression or leaflet calcification in patients who were receiving lipid-lowering medications than in control subjects and by experimental models showing that lipid-lowering therapy blocks the development of valve lesions.[7]
Thus, the results of the Simvastatin and Ezetimibe in Aortic Stenosis (SEAS) study (ClinicalTrials. gov number, NCT00092677) that are reported in this issue of the Journal8 are disappointing. In this large, randomized, prospective clinical trial, Rossebø et al. convincingly show that aggressive lipid lowering does not affect either hemodynamic progression or the time to valve replacement in adults with aortic stenosis.[8] Although it is possible that treatment even earlier in the disease process might have some benefit, the study patients
had only mild-to-moderate disease. Other than those with a bicuspid valve or specific genetic
markers, earlier identification of patients at risk would be problematic. The reduction in
atherosclerotic clinical end points in this study is encouraging. However, the clinical effect was
small, given that benefit was primarily due to a reduced rate of coronary bypass grafting at the
time of valve replacement.
If intensive lipid-lowering therapy is not the answer to the prevention of aortic stenosis progression, where do we go from here? Many adults with calcific valve disease meet current indications for lipid-lowering therapy, and the SEAS study certainly supports the evaluation and reduction of risk factors in patients with aortic stenosis, as recommended for all adults by established guidelines. However, we can no longer reassure ourselves that either the lipid-lowering or pleiotropic effects of potent agents such as statins and ezetimibe might change the disease process in the valve leaflets. We need to explore other potential therapeutic targets, especially the pathways that lead to tissue calcification. Calcific aortic stenosis is not atherosclerosis. Although there is overlap in clinical risk factors, in tissue characteristics, and in the association between the presence of calcific valve disease and atherosclerotic clinical events, there also are major differences. In aortic valve stenosis, tissue calcification is more severe; the mechanism of clinical events is increased leaflet stiffness, not plaque rupture; and the severity of coronary and valve disease in an individual patient often is discordant.
Demonstrating clinical benefit of potential therapies for calcific aortic stenosis will be challenging.
The evaluation of clinical end points requires a large study group, and enrollment in prospective, randomized trials is slow, given the relatively low prevalence of valve disease. Calcific valve disease progresses slowly over decades, whereas clinical trials usually follow patients for
only a few years. Clinical end points often are difficult to assess because indications for valve
replacement remain somewhat subjective and because therapy may affect other cardiovascular
end points, which limits our understanding of the mechanism of benefit. Doppler echocardiography allows assessment of the effect of therapy on the degree of stenosis, but it is not perfect, because hemodynamic obstruction occurs only with a substantial amount of leaflet thickening. The ideal end point for measuring the effect of therapy would be direct evaluation of tissue changes in the valve leaflets. Such analysis is possible in experimental models but in humans is limited to the examination of leaflets removed at the time of valve surgery.[9] Computed tomographic imaging allows measurement of leaflet calcification but not of other tissue components. In the future, molecular imaging approaches may provide sensitive measures of tissue changes sequentially over time, allowing detection of significant differences between small study groups.[10]
It is time to integrate and expand our understanding of the interactions between initiating
factors, genetic and clinical cofactors, and the mechanisms of progression from an early inflammatory lesion to phenotypic transformation of valve myofibroblasts and then to the end stage of severe valve calcification. Discovery of an effective medical therapy for calcific aortic stenosis will require innovative approaches to disease prevention and ingenuity in proving the mechanism of benefit.
From the Division of Cardiology, Department of Medicine, University
of Washington, Seattle.
1. Roberts WC, Ko JM. Frequency by decades of unicuspid, bicuspid,
and tricuspid aortic valves in adults having isolated aortic
valve replacement for aortic stenosis, with or without associated
aortic regurgitation. Circulation 2005;111:920-5.
2. Katz R, Wong ND, Kronmal R, et al. Features of the metabolic
syndrome and diabetes mellitus as predictors of aortic
valve calcification in the Multi-Ethnic Study of Atherosclerosis.
Circulation 2006;113:2113-9.
3. Garg V, Muth AN, Ransom JF, et al. Mutations in NOTCH1
cause aortic valve disease. Nature 2005;437:270-4.
4. Helske S, Oksjoki R, Lindstedt KA, et al. Complement system
is activated in stenotic aortic valves. Atherosclerosis 2008;
196:190-200.
5. Akat K, Borggrefe M, Kaden JJ. Aortic valve calcification —
basic science to clinical practice. Heart 2008 July 16 (Epub ahead
of print).
6. Weiss RM, Ohashi M, Miller JD, Young SG, Heistad DD. Calcific
aortic valve stenosis in old hypercholesterolemic mice. Circulation
2006;114:2065-9.
7. Rajamannan NM, Subramaniam M, Caira F, Stock SR, Spelsberg
TC. Atorvastatin inhibits hypercholesterolemia-induced
calcification in the aortic valves via the Lrp5 receptor pathway.
Circulation 2005;112:Suppl:I229-I234.
8. Rossebø AB, Pedersen TR, Boman K, et al. Intensive lipid
lowering with simvastatin and ezetimibe in aortic stenosis.
N Engl J Med 2008;359:1343-56.
9. Anger T, Pohle FK, Kandler L, et al. VAP-1, Eotaxin3 and
MIG as potential atherosclerotic triggers of severe calcified and
stenotic human aortic valves: effects of statins. Exp Mol Pathol
2007;83:435-42.
10. Aikawa E, Nahrendorf M, Sosnovik D, et al. Multimodality
molecular imaging identifies proteolytic and osteogenic activities
in early aortic valve disease. Circulation 2007;115:377- 86.

Diet, Exercise and the Metabolic Syndrome

Christos Pitsavos1, Demosthenes Panagiotakos2, Michael Weinem3 and Christodoulos Stefanadis1

1 First Cardiology Clinic, School of Medicine, University of Athens, Athens, Greece. 2 Department of Nutrition and Dietetics,
Harokopio University, Athens, Greece. 3 Society for Biomedical Diabetes Research, Duisburg, Germany. Address correspondence to: Demosthenes B. Panagiotakos, e-mail: d.b.panagiotakos@usa.net

■ Abstract
The metabolic syndrome is a combination of metabolic disorders, such as dyslipidemia, hypertension, impaired glucose tolerance, compensatory hyperinsulinemia and the tendency
to develop fat around the abdomen. Individuals with the metabolic syndrome are at high risk for atherosclerosis and, consequently, cardiovascular disease. However, as a result of several epidemiologic studies and some clinical trials, it has been suggested that people with the metabolic syndrome may benefit from intensive lifestyle modifications including dietary changes and adopting a physically more active lifestyle. In this review we summarize the effects of diet and physical activity on the development of the metabolic syndrome.
Keywords: metabolic syndrome · diet · exercise · lifestyle

Defining the metabolic syndrome
he metabolic syndrome is a collection of conditions associated with metabolic disorder and increased risk of developing cardiovascular disease. Conditions such as dyslipidemia, high blood pressure, impaired glucose tolerance and abdominal fat accumulation
fall into this category [1-5]. Investigations were aimed at the establishment of a quantitative definition for the associated conditions. However, these efforts led to multiple definitions that are partly inconsistent and disputed.
The metabolic syndrome was first described in the 1940s by Jean Vague, who linked abdominal obesity to metabolic abnormalities. Three decades later, in the 1970s, Gerald Phillips, suggested that aging, obesity and sex hormone-associated clinical manifestations, now referred to as the metabolic syndrome, are associated with heart disease [1]. More recently, in 1988, Gerald
Reaven proposed insulin resistance, and not obesity, as the critical factor and named the constellation of abnormalities Syndrome-X [2]. However, the most widely used definitions were established by the World Health Organization (WHO) and the National Cholesterol
Education Program Adult Treatment Panel III (NCEP ATPIII). These organizations regarded the metabolic syndrome as a cardiovascular risk factor beside elevated low-density lipoprotein (LDL) cholesterol [6, 7]. Atherogenic dyslipidemia (a prothrombotic state), insulin resistance, hypertension, abdominal obesity and elevated levels of various inflammatory markers were then regarded as the prominent characteristics of the metabolic syndrome [8]. Kim and Reaven
claimed that although WHO and NCEP ATPIII use the same term to define this condition (i.e., metabolic syndrome), both pursue different diagnostic aims and use different criteria to identify individuals, which relate to their different institutional goals [9]. In contrast to the WHO definition, the NCEP ATPIII does not include the measurement of insulin and therefore may
fail to detect insulin resistance. Nevertheless, the NCEP ATPIII definition appears to be more predictive regarding the risk of developing the syndrome than the WHO definition, i.e. failure to detect insulin resistance need not be a disadvantage.
In 2005, the International Diabetes Federation (IDF) Epidemiology Task Force suggested a new definition for the metabolic syndrome, focusing on central obesity [10]. Comparing the three definitions with respect to their areas of use, the WHO criteria appear to be more suitable for research purposes, while the NCEP ATPIII and IDF criteria seem to be more useful for clinical practice. The latter require only fasting assessments of blood samples, while WHO criteria require oral glucose tolerance tests, which can meaningfully confirm insulin resistance but are less practical in epidemiological or clinical studies. In general, the NCEP ATPIII and IDF definitions give more weight to obesity and sedentary lifestyle, whereas the WHO emphasizes the importance of insulin resistance as an underlying etiology of the metabolic syndrome. The
common feature of all three definitions is that the definition of the metabolic syndrome should include characteristics of atherogenic dyslipidemia, insulin resistance, hypertension and obesity. However, only the IDF definition considers obesity as a prerequisite and takes into account that obesity in Asian and other populations differs in its definition from obesity in Europeans [10]. Nevertheless, each of the defining abnormalities may promote atherosclerosis independently,
but when clustered together, these metabolic disorders, beside elevated LDL cholesterol, are increasingly atherogenic and may substantially enhance the risk of cardiovascular disease. Because each independent factor of the metabolic syndrome can increase the individual’s cardiovascular risk, an integrated and comprehensive approach is necessary for people afflicted
with the syndrome.
It is now widely accepted that the treatment of hypertension, obesity and dyslipidemia should be primarily based on weight-loss diets and exercise programs to increase physical activity and to ameliorate progress of the symptoms. In this review we present a summary and assessment of the existing research regarding interventions in the metabolic syndrome and of epidemiologic
studies on diet and exercise in relation to the prevalence of the metabolic syndrome (or diabetes, which may lead to the development of the syndrome).

Epidemiology of the metabolic syndrome
In this section, we review studies on lifestyle changes and metabolic syndrome. However, first we examine the prevalence of this condition at the population level. It is supposed that a substantial proportion of individuals living in Western nations are afflicted with multiple metabolic abnormalities [3]. A recently published report by the NCEP ATPIII estimates that
at least 47 million Americans are afflicted with this condition and projects the number of US citizens with metabolic syndrome to be between 50 to 75 million in 2010 [11]. Considering Europe, Hu et al. from the DECODE Study group reported that the agestandardized
prevalence of the metabolic syndrome was 15.7% in men and 14.2% in women [12]. For the
Mediterranean region, Ferrannini et al. estimated that more than 70% of adults have at least one of the major characteristics of the metabolic syndrome [13]. In this context, the ATTICA Study, comprising 1,500 women and men from Greece, estimated the prevalence of the
metabolic syndrome at 25% in men and 15% in women [14]. Recently, Athyros et al., who considered a Northern Greek population, reported that the ageadjusted prevalence of the NCEP ATP III-defined metabolic syndrome was 25% whereas the IDFdefined prevalence was 43% [15]. Furthermore, the prevalence of the metabolic syndrome in a Portuguese population was 27% in women and 19% in men [16]. A very similar result was derived from an examination
of a Korean population, where the prevalence of the metabolic syndrome was 29% in men and 17% in women [17], while in another study of the same population the prevalence of the syndrome was only 13% in both men and women [18]. Differences in genetic background, dietary habits, levels of physical activity, population age and sex structure and levels of overand
under-nutrition may influence the prevalence of both the metabolic syndrome and its components worldwide. Nevertheless, all these epidemiologic studies
suggest that the prevalence of the syndrome is high worldwide. This could be due to increasing obesity and sedentary lifestyles and reflect the growing necessity for therapeutic intervention.

The role of diet in the treatment of the metabolic syndrome
The NCEP ATPIII suggested therapeutic lifestyle changes (TLC) in order to reduce the prevalence of the metabolic syndrome [11]. Among several factors related to lifestyle habits the beneficial effect of diet has already been highlighted in many clinical and epidemiological studies [19-29]. During the last decades increasing scientific evidence has emerged that protective
health effects can be obtained from diets that are rich in fruits, vegetables, legumes and whole grains, and which include fish, nuts, and low-fat dairy products. Such diets need not be restricted in total fat in take as long as energy intake does not exceed caloric expenditure and if they emphasize predominantly vegetable oils that have a low content of saturated fats
and partially hydrogenated oils. As the intake of specific nutrients may have different effects on the development of metabolic syndrome characteristics the following sections focus on separate nutrient groups in order to clarify their roles in disease and treatment.

Nutrients and the metabolic syndrome
Carbohydrate consumption has been a critical factor blamed for weight gain, obesity, diabetes, and a number of other diseases. It is important to recognize that such problems may be associated with the excess consumption of the wrong carbohydrates such as simple sugars (i.e., table sugar), but not with complex carbohydrates. Large proportions of complex carbohydrates
(such as potatoes, breads, corn, etc.) in the diet are recommended.
High-fiber diets have received considerable attention in recent years due to their association with decreased incidence of several metabolic disorders such as hypertension, diabetes, obesity, as well as heart disease and colon cancer.
Fat is a general term used to refer to oils, fats and waxes. Usually the daily energy intake consists of 30% fat, but no more than 10% of these calories should come from saturated (animal) fats. The residual energy should be obtained from polyunsaturated or monounsaturated
oils [27]. Saturated fats promote dyslipidemias and, consequently atherogenesis. The consumption of unsaturated fats, derived mostly from vegetable oils such as safflower, corn,
olive and soybean oil, may be able to prevent serious disorders, such as atherogenesis,
hypertension and consequently the metabolic syndrome.
Nutritional studies suggest that we only need relatively small amounts of protein for
good health. The requirements for adults are 0.8 grams per kilogram of body weight. Increased protein intake may be detrimental for obese persons and those with kidney disease [30].

Dietary patterns
Diets should include a balanced intake of nutrient elements [27]. During the past two decades a large body of evidence has related balanced dietary patterns, such as the Mediterranean, to lower mortality rates, decreased prevalence of some metabolic disorders (obesity, high blood pressure), as well as lower incidence of coronary heart disease and various types of cancer.
The Mediterranean dietary pattern has received much attention in the last ten years [22, 24-29, 31]. It is characterized by the use of olive oil, which is important not only because it has several beneficial properties, but also because it allows the consumption of large quantities of vegetables in the form of salads and equally large quantities of legumes in the form of cooked foods. Other essential components of the Mediterranean diet are wheat, olives and grapes, and their various derivative products. Total lipid intake may be high - around or in excess of 40% of total energy intake - however, the ratio of monounsaturated to saturated fats is much higher in the Mediterranean regions than in other places of the world. A potential explanation for the beneficial effect of this dietary pattern on human health is that it is low in saturated fat,
high in monounsaturated fat, mainly from olive oil, high in complex carbohydrates from legumes, and high in fiber, mostly from vegetables and fruits. The high content of vegetables, fresh fruits, cereals and olive oil guarantees a high intake of beta-carotene, vitamins C and E, polyphenols and various important minerals. These key elements have been suggested to be responsible
for the beneficial effect of this diet on human health [22]. Interestingly, during the last years, several researchers have associated the Mediterranean diet with improvements in the blood lipid profile (in particular HDL cholesterol and oxidized LDL), decreased risk of thrombosis (i.e., fibrinogen levels), improvements in endothelial function and insulin resistance, reduction in plasma homocysteine concentrations, and a decrease in body fat [24-29, 31].
Furthermore, antioxidants represent a common element in the Mediterranean diet and antioxidant action provides a plausible explanation for its apparent benefits [27]. It is known that wild edible greens frequently eaten in the form of salads and pies contain very high quantities of flavonoids. Although there is no direct evidence that these antioxidants are central to the benefits of the Mediterranean diet, indirect evidence from epidemiological data and an increasing understanding of their mechanisms of action suggest that antioxidants may play a major role. Recently, the ATTICA Study investigators showed that adherence to
the Mediterranean diet was associated with 20% lower odds of having the metabolic syndrome, irrespective of age, sex, physical activity, lipids and blood pressure levels [14].

The role of exercise
In the late 1970s several observational studies suggested that mortality or morbidity caused by atherosclerotic disease was inversely related to the individual’s physical activity status [32-40]. Even though exercise is considered a cornerstone in the treatment of diabetes, a condition that is strongly related to metabolic syndrome, only a few studies have investigated its relationship with cardiovascular disease risk in diabetic persons. In a sample of 492 diabetic men and women
from the National Health and Nutrition Examination Survey, followed-up for 2 years, Ford and DeStefano [36] found that inactivity in non-leisure time was significantly associated with higher rates of coronary death. Data from an average 8.2-year, prospective, follow- up of 8,715 men in a preventive medicine clinic in the USA demonstrated a higher risk of all-cause mortality
for unfit compared to fit persons, within each of three glycemic status levels [37]. See Table 2 for a summary of studies evaluating physical activity in relation to the metabolic syndrome or associated conditions.
In a sample of 1,263 diabetic men, followed-up for 12 years in the Aerobics Center Longitudinal Study, participants who reported being sedentary had an adjusted risk for mortality of 1.7 compared to those who were physically active [38]. In another sample of 5,125 diabetic nurses from the Nurses Health Study, after 14 years of follow-up, the investigators found a 45% multivariate- adjusted reduction in cardiovascular disease risk with moderate to vigorous activity compared to sedentary [39]. The Whitehall Cohort Study investigated the relation of two indices of physical activity - walking pace and leisure activity - to total mortality,
coronary heart disease and other cardiovascular diseases, in a 25-years follow-up of 6,408 male British civil servants [40]. Among 352 diabetic men and 6,056 non-diabetics at study entry, the investigators found that the two indices of physical activity were inversely related to all-cause, coronary heart disease and other cardiovascular disease mortalities in both normoglycemic
men and men with diabetes/impaired glucose tolerance.
More recently, Tanasescu et al. [41] from the Health Professionals’ Study, during a 14-year follow-up of 2803 men, observed a 42% multivariate-adjusted reduction of total mortality and a 33% multivariateadjusted reduction of cardiovascular disease incidence in the highest quintile of physical activity compared with the lowest.
The Finish Diabetes Prevention Study (DPS), a randomized clinical trial including 522 men and
women with impaired glucose tolerance, intended to investigate if leisure-time physical activity is associated with the prevalence of type 2 diabetes [42]. The goal for physical activity in leisure times was an exercise of ≥ 30 min/day. The study showed that people with increased moderate-to-vigorous leisure time physical activity were 65% less likely to develop diabetes after various adjustments for changes in diet and body weight. In a similar study, the Diabetes Prevention
Program (DPP) included 3,234 obese subjects with impaired glucose tolerance but not diabetes and randomized them to metformin, lifestyle changes (diet and exercise) and placebo [43]. The investigators introduced a lifestyle-modification program with the goals of at least a 7 percent weight loss and a physical activity of ≥ 150 min/wk. It could be observed that both treatments, lifestyle changes and metformin, were significantly different to placebo. However, lifestyle
changes were more effective than metformin with a reduced incidence of diabetes of 58% (lifestyle) compared to 31% (metformin) [43].
In contrast to the number of studies that investigated the association of exercise with the development of diabetes or cardiovascular disease, data considering specifically the metabolic syndrome are sparse in the literature. One of the epidemiologic studies that evaluated
the association between physical activity and the prevalence of the metabolic syndrome was the ATTICA Study [14]. The results showed that even lightto- moderate leisure time physical activity (<7 kcal/min expended) was associated with a considerable reduction in the prevalence of the metabolic syndrome in 3042 men and women from the general population. Regular, intensive exercise was associated with a much greater decrease [14]. In addition, the ATTICA Study investigators demonstrated that the adoption of the Mediterranean diet by physically active people was associated with greater reduction in the odds of having the syndrome than diet or exercise alone, after adjusting for several potential confounders. Thus, the combination
of beneficial health factors in terms of nutrition and exercise explained at least a part of the reduction in the prevalence of the metabolic syndrome; and this effect still remained beneficial when considering differences in lipids as well as inflammation and coagulation factors [44].
The level of physical activity needed for a beneficial impact on coronary risk remains controversial. The Center for Disease Control and Prevention and the American College of Sports Medicine recommend the accumulation of at least 30 minutes of moderateintensity
physical activity (equivalent to brisk walking at 3-4 mph), on most, preferably all, days of the week on the basis of documented improvements in fitness, for the general population [45]. This level of activity is well tolerated by most middle-aged or older individuals. However, people who are initially unfit or sedentary should start at lower intensity. Nevertheless, it could be strongly suggested that even low levels of physical activity may modify the status of the clinical and biochemical components of the metabolic syndrome and, therefore reduce its prevalence in the
population [45].
The protective role of physical activity has been attributed to various mechanisms. On the one hand, physical exercise has favorable effects on traditional cardiovascular risk factors; on the other, the positive effect can be attributed to a direct action of physical activity on the heart itself leading to increased myocardial oxygen supply, decreased myocardial oxygen demands,
formation of collateral coronary circulation, improved myocardial contraction and electrical stability of the heart [45].
The theoretical mechanism for chronic exercise promoting a reduction in body fat involves increased total daily energy expenditure without a corresponding increase in energy intake. It is generally accepted that long-term physical activity of sufficient intensity, duration and frequency has a favorable effect on weight reduction and body fat distribution. Evidence supports the hypothesis that the effectiveness of exercise to induce weight loss is directly related to the initial degree of obesity and the total amount of energy expenditure [46].
The beneficial effect of physical activity on blood pressure levels has also been shown [47-49]. In particular, it is now accepted that moderate levels of exercise can significantly decrease blood pressure in patients with mild to moderate essential hypertension.
Although physical activity has an insignificant effect on blood lipid levels, some investigators have shown the overall benefit of physical activity in modifying blood lipid profiles. The Pawtucket Heart Study group reported that physical activity was significantly associated
with higher HDL-cholesterol levels [49]. Moreover, among 3,000 adult Japanese men the frequency of physical activity was independently and positively related to HDL-cholesterol [50]. Similarly, a pooled analysis among three European cohorts consisting of elderly men demonstrated a significant relation between physical activity and HDL-cholesterol [51]. Reports
by Ford [52], and King [53] studying approximately 14,000 adult participants in the National Health and Nutrition Examination Survey III (1988-1994) showed that the time devoted to physical activity was inversely associated with some inflammatory marker levels, such as C-reactive protein, plasma fibrinogen concentration and the number of white blood cells, after
adjusting for several potential confounders. Similarly, Abramson et al. [54] reported that physical activity was independently associated with a lower probability of having elevated inflammatory marker levels among healthy US adults aged 40 years and older, independent
of several confounding factors. An inverse relation between plasma fibrinogen levels and leisure time physical activity has also been reported by several others [55-58].
Lifestyle approaches to treating and preventing the metabolic syndrome vary, but nearly all experts agree that parameters involved in the syndrome are greatly improved by reducing body weight and increasing the level of physical activity . Recently, Roberts et al. [59] and Stone et al. [60] revealed by an extensive review of the literature that lifestyle modifications mitigated disease progression and reversed existing disease. Small changes can lead to great improvements, not for achieving a perfect lifestyle but for working towards a better and healthier one. However, it should be noted that although lifestyle changes can provide
many benefits for human health, and especially for the management of the metabolic syndrome, sometimes these changes are difficult to implement and maintain. Therefore, drug treatment including statins, ACE inhibitors, angiotensin-II receptor blockers, and oral antidiabetic agents can be considered. It has been shown that these drugs are able to reduce effectively
the levels of underlying risk factors for the metabolic syndrome such as dyslipidemia, hypertension, hyperglycemia and the risk of developing diabetes [61].

Concluding remarks
The metabolic syndrome seems to be an emerging epidemic that affects roughly one out of five persons in Western industrialized countries. Similar to other chronic diseases, the metabolic syndrome is a complex, lifestyle-dependent illness. Its solution is not difficult to achieve: eat less, exercise more. These solutions must become part of everyday life and be woven into our social life to be effective. Health care professionals need to help people to understand the potential benefits that may result from the introduction of dietary patterns and exercise, and support them in adopting and adhering to these behavioral patterns. Actually, society as a whole needs to acquire a profound consciousness of the relevance for health of lifestyle factors such as nutrition and activity.

References
1. Phillips GB. Sex hormones, risk factors and cardiovascular
disease. Am J Med 1978. 65:7-11.
2. Reaven GM. Role of insulin resistance in human disease. Diabetes
1988. 37:1595-607.
3. Grundy SM. Obesity, metabolic syndrome and cardiovascular
disease. J Clin Endocrinol Metab 2004. 89:2595-600.
4. Ford ES, Giles WH, Dietz WH. Prevalence of the metabolic
syndrome among US adults: findings from the third National
Health and Nutrition Examination Survey. JAMA 2002.
287:356-359.
5. Das UN. Metabolic syndrome X: an inflammatory condition?
Curr Hypertens Rep 2004. 6:66-73.
6. Alberti KG, Zimmet P. Definition, diagnosis and classification
of diabetes mellitus and its complications. Part 1: diagnosis
and classification of diabetes mellitus provisional report of a
WHO consultation. Diabet Med 1998. 15:539-53.
7. Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel
RH, Franklin BA, Gordon DJ, Krauss RM, Savage PJ,
Smith SC Jr, Spertus JA, Costa F. Diagnosis and Management
of the Metabolic Syndrome. An American Heart Association/
National Heart, Lung, and Blood Institute Scientific
Statement. Circulation 2005. 112(17):2735-2752.
8. Haffner SM. The metabolic syndrome: inflammation, diabetes
mellitus, and cardiovascular disease. Am J Cardiol 2006.
97(2A):3A-11A.
9. Kim SH, Reaven GM. The metabolic syndrome: one step
forward, two steps back. Diab Vasc Dis Res 2004. 1:68-75.
10. Alberti KG, Zimmet P, Shaw J; IDF Epidemiology Task
Force Consensus Group. The metabolic syndrome - a new
worldwide definition. Lancet 2005. 366:1059-1062.
11. 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. 285:2486-
2497.
12. Hu G, Qiao Q, Tuomilehto J, Balkau B, Borch-Johnsen
K, Pyorala K; DECODE Study Group. Prevalence of the
metabolic syndrome and its relation to all-cause and cardiovascular
mortality in non-diabetic European men and women.
Arch Intern Med 2004. 164:1066-1076.
13. Ferrannini E, Natali A. Essential hypertension, metabolic
disorders, and insulin resistance. Am Heart J 1991. 121:1274-
1282.
14. Panagiotakos DB, Pitsavos CH, Chrysohoou C, Skoumas
J, Tousoulis D, Toutouza M, Toutouzas PK, Stefanadis C.
The Impact of Lifestyle Habits on the Prevalence of the Metabolic
Syndrome among Greek adults from the ATTICA study.
Am Heart J 2004. 147:106-112.
15. Athyros VG, Ganotakis ES, Elisaf M, Mikhailidis DP. The
prevalence of the metabolic syndrome using the National Cholesterol
Educational Program and International Diabetes Federation
definitions. Curr Med Res Opin 2005. 21:1157-1159.
16. Santos AC, Lopes C, Barros H. Prevalence of metabolic
syndrome in the city of Porto. Rev Port Cardiol 2004. 23:45-52.
17. Oh JY, Hong YS, Sung YA, Barrett-Connor E. Prevalence
and factor analysis of metabolic syndrome in an urban Korean
population. Diabetes Care 2004. 27:2027-2032.
18. Lee WY, Park JS, Noh SY, Rhee EJ, Kim SW, Zimmet PZ.
Prevalence of the metabolic syndrome among 40,698 Korean
metropolitan subjects. Diabetes Res Clin Pract 2004. 65:143-149.
19. Kannel WB, McGee DL, Gordon T. A general cardiovascular
risk profile: The Framingham Study. Am J Cardiol 1976.
38:46-51.
20. Keys A, Menotti A, Karvonen MJ, Aravanis C, Blackburn
H, Buzina R, Djordjevic BS, Dontas AS, Fidanza F, Keys
MH, et al. The diet and 15-year death rate in the Seven Countries
Study. Am J Epidemiol 1986. 124:903-915.
21. Trichopoulou A, Kouris-Blazos A, Wahlqvist M,
Trichopoulos D. Diet and overall survival in elderly people.
Br Med J 1995. 311:1457-1460.22. Trichopoulou A., Lagiou P. Healthy traditional Mediterranean
diet - An expression of culture, history and lifestyle. Nutr
Rev 1997. 55:383-389.
23. Knapp HW. Dietary fatty acids in human thrombosis and
homeostasis. Am J Clin Nutr 1997. 65:1687S-1698S.
24. de Lorgeril M, Salen P, Martin JL, Monjaud I, Delaye J,
Mamelle N. Mediterranean diet, traditional risk factors, and
the rate of cardiovascular complications after myocardial infarction:
final report of the Lyon Diet Heart Study. Circulation.
1999;99:779-85.
25. Panagiotakos DB, Pitsavos C, Chrysohoou C, Tzioumis
K, Papaioannou I, Stefanadis C, Toutouzas PK. The Association
of Mediterranean Diet with Lower Risk of Acute Coronary
Syndromes, in Hypertensive Subjects. Int J Cardiol 2002.
19:141-147.
26. Martinez-Gonzalez MA, Fernandez-Jarne E, Serrano-
Martinez M. Mediterranean diet and reduction in the risk of a
first acute myocardial infarction: an operational healthy dietary
score. Eur J Nutr 2002. 41:153-160.
27. World Health Organization Study Group. Diet, Nutrition,
and the Prevention of Chronic Diseases. Geneva, Switzerland,
World Health Organization, Technical Report Series, 916, 2003.
28. Trichopoulou A, Costacou T, Bamia C, Trichopoulos D.
Adherence to a Mediterranean diet and survival in a Greek
population. N Engl J Med 2003. 348:2599-2608.
29. Chrysohoou C, Panagiotakos DB, Pitsavos C, Das UN,
Stefanadis C. Adherence to the Mediterranean diet attenuates
inflammation and coagulation process, in healthy adults: the
ATTICA study. J Am Col Cardiol 2004. 44:152-158.
30. Anderson CA, Miller ER 3rd. Dietary recommendations for
obese patients with chronic kidney disease. Adv Chronic Kidney
Dis 2006. 13(4):394-402.
31. Schroder H, Marrugat J, Vila J, Covas MI, Elosua R. Adherence
to the traditional Mediterranean diet is inversely associated
with body mass index and obesity in a Spanish population.
J Nutr 2004. 134:3355-3361.
32. American College of Sports Medicine. Position statement
on proper and improper weight loss programs. 1983. 15:9-13.
33. Leon A S, Connett J, Jacobs DR. Leisure-time physical activity
levels and risk of coronary heart disease and death: the Multiple
Risk Factor Intervention Trial. JAMA 1987. 258:2388-
2395.
34. Pekkanen J, Marti B, Nissinen A. Reduction of premature
mortality by high physical activity: a 20-year follow-up of middle-
aged Finnish men. Lancet 1987. 255:1473-1477.
35. Pate RR, Pratt M, Blair SN, Haskell WL. Physical activity
and public health. JAMA 1995. 273:402-407.
36. Ford ES, DeStefano F. Risk factors for mortality from all
causes and from coronary heart disease among persons with
diabetes. Findings from the National Health and Nutrition Examination
Survey I Epidemiologic Follow-up Study. Am J Epidemiol
1991. 133:1220-1230.
37. Kohl HW, Villegas JA, Gordon NF, Blair SN. Cardiorespiratory
fitness, glycemic status and mortality risk in men. Diabetes
Care 1992. 15:184-192.
38. Wei M, Gibbons LW, Kampert JB. Low cardiorespiratory
fitness and physical inactivity as predictors of mortality in men
with type 2 diabetes. Ann Intern Med 2000. 132:605-611.
39. Hu FB, Stampfer MJ, Solomon C. Physical activity and risks
for cardiovascular events in diabetic women. Ann Intern Med
2001. 134:96-105.
40. Batty GD, Shipley MJ, Marmot M, Smith GD. Physical activity
and cause-specific mortality in men with type 2 diabetes/
impaired glucose tolerance: evidence from the Whitehall
study. Diabet Med 2002. 19:580-588.
41. Tanasescu M, Leitzmann MF, Rimm EB, Hu FB. Physical
activity in relation to cardiovascular disease and total mortality
among men with type 2 diabetes. Circulation 2003. 107:2435-
2439.
42. Tuomilehto J, Lindstrom J, Eriksson JG, Valle TT,
Hamalainen H, Ilanne-Parikka P, Keinanen-
Kiukaanniemi S, Laakso M, Louheranta A, Rastas M,
Salminen V, Uusitupa M, Finnish Diabetes Prevention
Study Group. Prevention of type 2 diabetes mellitus by
changes in lifestyle among subjects with impaired glucose tolerance.
N Engl J Med 2001. 344(18):1343-1350.
43. Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF,
Lachin JM, Walker EA, Nathan DM, Diabetes Prevention
Program Research Group. Reduction in the incidence of
type 2 diabetes with lifestyle intervention or metformin. N Eng
J Med 2002. 346(6):393-403.
44. Pitsavos C, Panagiotakos DB, Chrysohoou C, Kavouras S,
Stefanadis C. The associations between physical activity, inflammation,
and coagulation markers, in people with metabolic
syndrome: the ATTICA study. Eur J Cardiovasc Prev Rehabil
2005. 12:151-158.
45. US Department of Health and Human Services. Physical
Activity and Health: a Report of the Surgeon General. Atlanta,
GA, US Department of Health and Human Services, Centers
for Disease Control and Prevention, National Center for
Chronic Disease Prevention and Health Promotion, 1996.
46. Jakicic JM, Winters C, Lang W, Wing RR. Effects of intermittent
exercise and use of home exercise equipment on adherence,
weight loss, and fitness in overweight women. JAMA
1999. 282:1554-1560.
47. Kokkinos PF, Narayan P, Colleran JA, Pittaras A, Notargiacomo
L, Reda D, Papademetriou V. Effects Of Regular
Exercise On Blood Pressure And Left Ventricular Hypertrophy
In African-American Men With Severe Hypertension.
N Engl J Med 1995. 333:1462-1467.
48. Burt VL, Whelton P, Roccella EJ. Prevalence of hypertension
in the US adult population. Results from the Third National
Health and Nutrition Examination Survey, 1988-1991.
Hypertension 1995. 25(3):305-313.
49. Eaton CB, Lapane KL, Garber CE. Physical activity physical
fitness and coronary heart disease risk factors. Med Sci Sports
Exerc 1995. 27:340-346.
50. Hsieh SD, Yoshinaga H, Muto T, Sakurai Y. Regular
physical activity and coronary risk factors in Japanese men. Circulation
1998. 97:661-665.
51. Bijnen FC, Feskens EJ, Caspersen CJ. Physical activity and
cardiovascular risk factors among elderly men in Finland, Italy
and the Netherlands. Am J Epidemiol 1996. 143:553-561.
52. Ford ES. Does exercise reduce inflammation? Physical activity
and C-reactive protein among U.S. adults. Epidemiology 2002.
13:561-568.
53. King DE, Carek P, Mainous AG 3rd, Pearson WS. Inflammatory
Markers and Exercise: Differences Related to Exercise
Type. Med Sci Sports Exerc 2003. 35:575-581.
54. Abramson JL, Vaccarino V. Relationship between physical
activity and inflammation among apparently healthy middleaged
and older US adults. Arch Intern Med 2002. 162:1286-1292.
55. MacAuley D, McCrum EE, Stott G, Evans AE,
McRoberts B, Boreham CA, Sweeney K, Trinick TR.Physical activity, physical fitness, blood pressure, and fibrinogen
in the Northern Ireland health and activity survey. J Epidemiol
Community Health 1996. 50:258-263.
56. Danesh J, Collins R, Appleby P. Association of fibrinogen,
C-reactive protein, albumin, or leukocyte count with coronary
heart disease: meta-analyses of prospective studies. JAMA
1998. 279:1477-1482.
57. Panagiotakos DB, Pitsavos C, Chrysohoou C, Kavouras S,
Stefanadis C. The associations between leisure-time physical
activity and inflammatory and coagulation markers related to
cardiovascular disease: the ATTICA Study. Prev Med 2004.
40(4):432-437.
58. Koh KK. Inflammatory markers and the metabolic syndrome. J
Am Coll Cardiol 2005. 46:1978-1985.
59. Roberts CK, Barnard RJ. Effects of exercise and diet on
chronic disease. J Appl Physiol 2005. 98:3-30.
60. Stone N, Saxon D. Approach to treatment of the patient with
metabolic syndrome. Lifestyle therapy. Am J Cardiol 2005.
22:15E-21E.
61. Haffner S. Diabetes and the metabolic syndrome. When is it
best to intervene to prevent? Atherosclerosis 2006. 7:S3-S10.
62. Eriksson KF, Lindgarde F. No excess 12-year mortality in
men with impaired glucose tolerance who participated in the
Malmo Preventive Trial with diet and exercise. Diabetologia
1998. 41(9):1010-1016.
63. Pan XR, Li GW, Hu YH, Wang JX, Yang WY, An ZX, Hu
ZX, Lin J, Xiao JZ, Cao HB, et al. Effects of diet and exercise
in preventing NIDDM in people with impaired glucose
tolerance. The Da Qing IGT and Diabetes Study. Diabetes Care
1997. 20(4):537-544.
64. Wein P, Beischer N, Harris C, Permezel M. A trial of simple
versus intensified dietary modification for prevention of
progression to diabetes mellitus in women with impaired glucose
tolerance. Aust N Z J Obstet Gynaecol 1999. 39(2):162-166.

Relative risk of diabetes, dyslipidaemia, hypertension and the
metabolic syndrome in people with severe mental illnesses:
Systematic review and metaanalysis

David PJ Osborn*1,2, Christine A Wright1, Gus Levy1, Michael B King1,2,
Raman Deo1,2 and Irwin Nazareth3,4

Address: 1Department of Mental Health Sciences, (Royal Free Campus), University College London Medical School, Rowland Hill Street, London,
NW3 2PF, UK, 2Camden and Islington Mental Health and Social Care Trust, St Pancras Way, London, NW1 OPE, UK, 3Department of Primary
Care and Population Health, (Royal Free Campus) University College Medical School, Rowland Hill Street, London, NW3 2PF, UK and 4MRC
General Practice Research Framework, 158-160 North Gower Street, London NW1 2ND, UK
Email: David PJ Osborn* - d.osborn@medsch.ucl.ac.uk; Christine A Wright - christine.wright@pms.ac.uk; Gus Levy - gus_levy@hotmail.com;
Michael B King - m.king@medsch.ucl.ac.uk; Raman Deo - ramandeo@hotmail.com; Irwin Nazareth - i.nazareth@pcps.ucl.ac.uk
* Corresponding author

Abstract
Background: Severe mental illnesses (SMI) may be independently associated with cardiovascular risk factors and the metabolic syndrome. We aimed to systematically assess studies that compared diabetes, dyslipidaemia, hypertension and metabolic syndrome in people with and without SMI.
Methods: We systematically searched MEDLINE, EMBASE, CINAHL & PsycINFO. We hand
searched reference lists of key articles. We employed three search main themes: SMI,
cardiovascular disease, and each cardiovascular risk factor. We selected cross-sectional, case
control, cohort or intervention studies comparing one or more risk factor in both SMI and a
reference group. We excluded studies without any reference group. We extracted data on: study design, cardiovascular risk factor(s) and their measurement, diagnosis of SMI, study setting,
sampling method, nature of comparison group and data on key risk factors.
Results: Of 14592 citations, 134 papers met criteria and 36 were finally included. 26 reported on diabetes, 12 hypertension, 11 dyslipidaemia, and 4 metabolic syndrome. Most studies were cross sectional, small and several lacked comparison data suitable for extraction. Meta-analysis was possible for diabetes, cholesterol and hypertension; revealing a pooled risk ratio of 1.70 (1.21 to 2.37) for diabetes and 1.11 (0.91 to 1.35) of hypertension. Restricting SMI to schizophreniform illnesses yielded a pooled risk ratio for diabetes of 1.87 (1.68 to 2.09). Total cholesterol was not higher in people with SMI (Standardized Mean Difference -0.10 (-0.55 to 0.36)) and there were inconsistent data on HDL, LDL and triglycerides with some, but not all, reporting lower levels of HDL cholesterol and raised triglyceride levels. Metabolic syndrome appeared more common in SMI.
Conclusion: Diabetes (but not hypertension) is more common in SMI. Data on other risk factors were limited by poor quality or inconsistent research findings, but a small number of studies show greater prevalence of the metabolic syndrome in SMI.

Background
People with severe mental illness (SMI) such as schizophrenia and bipolar affective disorder are at greater risk of coronary heart disease (CHD) than people without such diagnoses [1-3]. The mutable risk factors for CHD are smoking, hypertension, diabetes mellitus and high ratio of total cholesterol to High Density Lipoprotein (HDL) cholesterol. Although, many people with SMI are likely to be heavy smokers, and less likely to succeed in smoking cessation [4], the relationship between SMI and CHD mortality is not wholly explained by smoking[3] and there has been increasing interest in the prevalence of diabetes and dyslipidaemia in people with SMI. Second generation antipsychotics may exacerbate features of the metabolic syndrome including abnormal glucose and lipid profiles [2,5,6]. But recent reviews have suggested that people with SMI are at risk of the metabolic syndrome including diabetes irrespective of antipsychotic
therapy [7,8]. People with SMI share other risk factors including unhealthy lifestyles [9] obesity and positive family histories [10].
We hypothesised that there were differences in the risk of abnormal glucose, blood pressure or lipid abnormalities between people with and without SMI. We searched for studies comparing the risk of diabetes or hyperglycaemia, hypertension, dyslipidaemia or a combination of these
factors (as components of the metabolic syndrome or as an overall CHD risk score). We did not aim to assess smoking since a systematic review has recently been published [4] and the conclusions are uncontroversial.

Methods
We searched for studies of diabetes or hyperglycaemia, hypertension, dyslipidaemia or combinations of these factors in people with and without SMI and systematically
reviewed the literature to appraise the epidemiological evidence. We estimated the strength of any association between SMI and these CHD risk factors.
Data sources and search strategy
We electronically searched MEDLINE, EMBASE, CINAHL, the Cochrane Library database & PsycINFO for articles in English, French, German, Italian or Spanish and sought papers published between 1897 and 2005 inclusively. We hand searched reference lists of review papers and made contact with authors and researchers to ensure comprehensive coverage. We piloted and modified our search strategy to retrieve all key papers in this field. The most
sensitive search included three broad search themes namely 1) Terms related to SMI, 2) cardiovascular diseases and 3) the risk factors of diabetes, lipid disorders, hypertension, the metabolic syndrome and cardiovascular risk scores. Synonym lists were constructed for each theme and the databases were searched using these synonyms as both thesaurus and free-text terms For SMI, we included all terms relating to psychotic disorders, schizophreniform disorders, bipolar affective disorders and psychotic depression. Similarly all synonyms for
search themes 2 and 3 were employed. We included an additional wider term for all mental disorders in a final search combined with both search themes 2 and 3. A combination
of these two approaches provided the most reliable results.
Study selection
We included cross sectional, case-control, cohort and intervention studies in which the risk factors of interest were available in a group with SMI and a reference group without SMI. We excluded pharmacological studies comparing CHD risk factors between different antipsychotics
and without any comparison data from people not prescribed these drugs as these studies could not shed light on comparative risk between people with and without SMI. We included all studies involving representative groups with SMI and noted whether they were sampled
from the community; outpatient settings, inpatients or from long stay psychiatric accommodation.
Screening process
Two or more authors independently read all titles and available abstracts to identify potentially relevant articles. Decisions were compared and disagreements were discussed
at steering group meetings involving all authors. We translated non-English articles to determine their relevance.
Data extraction
We extracted data on the type of study design, the setting and the source of the groups with and without SMI. We recorded the type and method of SMI diagnosis and the reported response rate. We extracted which CHD risk factors (e.g. diabetes) were reported and how they were
measured and defined. We noted whether all participants were screened for CHD risk or whether the outcome (e.g. diabetes) relied on screening and diagnosis being made during routine clinical care. Summary data (i.e. raw numbers and percentages) on the prevalence of risk factors were obtained for each group including raw numbers and percentages. Comparative statistics were noted including absolute differences in continuous outcomes or proportions
and estimates of relative risk such as odds ratios. Adjustment of main results for confounders was also noted.
Data synthesis
We defined three levels of evidence. The highest level were studies where a non-SMI comparison group was recruited. The next level were those that did not recruit a comparison
group but used comparative risk factors data from general health population studies and the lowest level included studies where a selected group with other psychiatric diagnoses was used as a comparison. Within these levels we then grouped studies according to SMI diagnosis,
and the sampling frame for the SMI group(e.g. community or from a specific secondary care setting such as an inpatient unit or clinic. Finally, where possible we calculated summary statistics such as risk ratios (RRs), confidence intervals and standardized mean differences (the
mean difference in outcome/standard deviation for outcome; the effect size) for outcomes even when the papers had not presented such results. This was only possible when papers either reported raw numbers for dichotomous outcomes or means plus standard deviations for
continuous outcomes.
Meta-analysis
Data were entered into Stata version 9 [11] and standard meta-analytic techniques were employed if there were more than three studies for a given outcome. Meta-analyses
could only conducted on studies that reported data from a comparison group. We calculated pooled estimates of effect sizes and risk ratios using a random effects model that uses inverse variance methods to apportion more weight to larger rather than smaller studies in the metaanalysis. We approached heterogeneity in results between studies in two ways. Firstly we assessed whether a significant level of difference existed using Mantel-Haenszel chi square tests. If the chi square test was significant below p = 0.05, we quantified the amount of heterogeneity using I2 statistics. We considered I2 above 50% as an indicative of substantial heterogeneity.
Where studies only reported percentages, we could only calculate risk ratios rather than odds ratios. For consistency we therefore present risk ratios rater than odds ratios
in the meta-analyses.

Results
The initial database search generated 14592 papers, 134 papers were identified for further scrutiny but more detailed assessment by up to four authors yielded 36 papers [12-47] that were eligible for inclusion in the final review.
27 papers reported outcomes related to diabetes or hyperglycemia, 14 reported hypertension or blood pressure, 12 dyslipidaemia or lipid levels, 5 the metabolic syndrome and 4 papers included overall cardiovascular risk scores such as a ten year Framingham risk score [48]. Of the 98
excluded papers, the most common reason for ineligibility was that the paper only explored specific comparisons between different antipsychotic drugs, without comparison data on people not taking the drugs (n = 23). Many studies reported cardiovascular outcomes in samples of people with SMI without any reference data (n = 38). In several studies, samples included people with multiple diagnoses other than our definition of SMI (such as dementia), with no specific data for the subgroups with SMI as defined in this paper.
Study characteristics
Of the 36 papers included, most (28/36) reported cross sectional data on one or more cardiovascular risk factors. There were 8 papers utilizing data from longitudinal studies
[14-16,18,20,35,39,44], although strictly, none collected consecutive longitudinal information regarding cardiovascular risk factors at both baseline and follow-up. Four studies [12,13,33,45] recruited specifically from community settings, eight studies [14-16,24,25,37,46,47] from a mixture of outpatient and inpatient settings, and 5 studies [22,23,34,42,43] from outpatient clinics. The remaining 19 studies collected data from acute or long stay inpatient samples. These papers are summarized in additional file 2; tables 1–5, by outcome of diabetes/ hyperglycaemia (additional file 2; table 1), hypertension/ blood pressure , dyslipidaemia/ lipid levels (additional file 2; table 3), metabolic syndrome (additional file 2; table 4) and 10 year CHD risk scores (additional file 2; table 5). Within these tables, the studies are grouped according to whether they recruited a comparison group or simply used general population data, and by source of the SMI sample (eg inpatients or community).
Diabetes and hyperglycaemia
Twenty seven eligible papers [12-36,46,47] reported glucose related outcomes . The
location of the studies, source, definition, SMI diagnosis and relevant outcomes for each study are summarized . Diabetes or hyperglycemia definitions included diagnosis or treatment for diabetes in the clinical records (n = 17) [12-18,22,24-27,30,34-36,46] self-reported diabetic diagnosis (n = 2) [23,25], screening results for random glucose (n = 2)[13,29] or fasting glucose(
n = 3) [20,28,47] and impaired glucose tolerance (n = 3) [19,31,32]. Most studies (n = 23) included people with diagnoses of schizophrenia and/or schizoaffective disorder, two studies also included people with bipolar affective disorder [27,46]. Three studies only included people with a diagnosis of bipolar disorder [26,30,35].
Nine studies provided data that could be used in the metaanalysis for diabetes [12-16,18-20,23] .
These studies involved 9612 people with SMI, 1166 of whom had a diagnosis of diabetes and a total of 3449677 people without SMI of whom 534248 had recognized diabetes. The pooled risk ratio for diabetes in SMI was 1.70 (1.21 to 2.37). There was considerable heterogeneity with
a significant overall test for heterogeneity (chi square = 57.91 p < 0.001; I2 = 91.2%). However, within the schizophrenia and/or schizoaffective disorder group there was no significant heterogeneity (p = 0.837, I2 < 0.1%). In this group risk ratios for recorded diabetes ranged from 0.54 to 9.0 and the pooled risk ratio was 1.87 (1.68 to 2.09). The derived risk ratio from the one study including participants with bipolar affective disorder was 1.10 (1.03 to 1.18). There was no significant difference in results from studies of inpatient SMI samples compared to community
samples (test for heterogeneity between subgroups p = 0.851).
displays risk ratios for diabetes in SMI and, where possible, confidence intervals . These are based on studies in which the data were unsuitable for inclusion in meta-analysis e.g.comparison data only reported as percentages or using general population statistics without raw data).
Four studies compared random[13] or fasting[19,20,47] glucose levels, of which two [13,20] showed significantly increased standardized mean differences in the SMI group. The CATIE study [47] reported mixed results which differed by gender. Mean fasting glucose was significantly raised in SMI females but not males. However males with SMI were significantly more likely to reach criteria for raised fasting glucose than controls but this finding was not repeated in females.
Hypertension
Fifteen papers reported data relating to hypertension (additional file 2; table 2). Most included people with schizophrenia or schizoaffective disorder but three papers
included people with bipolar affective disorder P[16,38,46]. Hypertension was variously assessed by: selfreport, existing use of anti-hypertensive medication, diagnosis of hypertension in clinical records, direct measurement of blood pressure and use of varying systolic and
diastolic thresholds for hypertension.
Seven studies that included 2333/6249 people with SMI who were hypertensive and 1261228/2169371 hypertensive people without SMI were included in the meta-analysis
and none showed significantly elevated risk for hypertension in the SMI group (figure 4). The pooled risk ratio for hypertension in SMI was 1.11 (0.91 to 1.35). Heterogeneity
was significant (Chi square test p < 0.001 I2 = 89.2%).
We were able to calculate risk ratios (but not confidence intervals) for hypertension from four further studies which also reported general population comparison data, resulting in two raised RRs [23,38] and two reduced RRs [34,46] for hypertension in SMI.
Dyslipidaemia
12 studies reported a variety of lipid outcomes including total cholesterol, High Density Lipoprotein (HDL) cholesterol and Low Density Lipoprotein (LDL) cholesterol and triglycerides (additional file 2; table 3), some using fasting samples and some not. Seven included raw data from a comparison group, and only three studies included people with bipolar affective disorder. The only lipid outcome reported in sufficient studies for meta-analysis was mean total cholesterol (figure 5). Total cholesterol values were available for 160 people with SMI and 5702 people without SMI in 4 different studies [13,19,39,40] The pooled SMD was -0.10 (0.55 to 0.36) (figure 5). There was significant heterogeneity; (chi square = 14.91, p = 0.002;I2 = 79.9%) with one study showing an SMI group with lower total mean cholesterol and another showing SMI the opposite result.
Standardized mean differences between SMI and non-SMI samples could be calculated for HDL cholesterol, LDL cholesterol and triglycerides in two studies [13,19].
One of these studies found significantly lower HDL levels and higher triglyceride levels [13]. The other found significantly lower LDL levels [19].
Some studies also reported lipid results based on: a diagnosis of hyperlipidaemia or lipid disorders (by ICD criteria) [14,16] receipt of antilipemic medication[14], or proportions of people with lipid levels exceeding a defined threshold [13,20,47]. These results were inconsistent.
In some studies people with SMI were significantly more likely to have low HDLP[13,47] high triglycerides[ 20,47] or a high HDL/total cholesterol ratio but in others these did not reach significance for either HDL[20] and LDL levels[13] or total cholesterol [13]. A calculated
risk ratio for prevalence of ICD 9 dyslipidaemic disorders was not significant (0.86: 0.73 to 1.01) [14].
Studies with insufficient data for calculating SMDs, or where other patient groups were used for comparison, also reported conflicting results. SMI was associated with significantly lower total cholesterol in one study[36] with significantly lower HDL cholesterol in others [37,47] but
this was not confirmed elsewhere[39]. Triglyceride results were also inconsistent [39,41,47].
Metabolic syndrome
Five papers involving 1026 people with schizophrenia or schizoaffective disorder reported prevalence of the metabolic syndrome according to international criteria, but only two studies used raw data from a comparison group in addition to 718 people with schizophrenia [20,47] . The other three studies used general population comparison data that were not suitable for meta-analysis (due to a lack of raw numbers) and not clearly age matched to the people with SMI. displays risk ratios (and where possible calculated confidence intervals) from all five studies. All point estimates for risk of the metabolic syndrome were raised.
Ten year cardiovascular risk scores
Four studies report ten year cardiovascular risk scores for people with SMI, two included comparison groups [13,44], and two used general population comparison data [37,45] (additional file 2; table 5). In two studies, involving 352 people with SMI, the 10 year cardiovascular risk scores was significantly increased in men but not women with SMI [37,45]. One study of 21 first-onset cases of schizophreniform illnesses showed significantly raised 10 year risk scores compared to general population, but no differences compared to matched controls [44]. Finally, one controlled community study found that excess cardiovascular risk scores were only detectable when different effects were considered at different age groups [13].

Discussion
We found that diabetes mellitus is the cardiovascular risk factor most convincingly associated with SMI. Meta-analysis of the highest quality studies revealed almost a two-fold risk of diabetes in schizophrenia-like illnesses but not bipolar affective disorder. Conversely, meta-analysis revealed no association between SMI and hypertension. Similar findings were observed for total cholesterol levels, but these studies were limited by their design and so conclusions
must be guarded. There were inadequate numbers of comparative studies of other lipids, such as HDL cholesterol, or of the metabolic syndrome to conduct a meta-analyses. Lower HDL cholesterol levels in people with SMI found in two studies [13,47] were not confirmed by another [20]. Five studies on the metabolic syndrome revealed an excess risk in people with SMI, and two, involving 718 people with SMI, confirmed such an excess of metabolic syndrome statistically. Raised "Framingham" or ten year cardiovascular risk scores may only be demonstrable in SMI when differences in effects are examined separately in different age
groups and sexes. For instance excess risk scores may only be detectable in those over 40.
Quality and variability of published studies
There were very few high quality comparative studies on cardiovascular risk factors and the metabolic syndrome in people with and without SMI. Many studies in this review
were limited by small, convenience samples of people with SMI such as those in specific clinics or inpatient units, compromising the generalisability of their findings. Furthermore, several studies reporting higher levels of cardiovascular risk in SMI did not obtain raw comparison
data within their study to allow statistical assessment of the importance of their findings. Cardiovascular risk factors are increasing rapidly in the general population, hence the need for relevant contemporary comparison figures. There were no studies designed to compare the longitudinal development of cardiovascular risk factors between people with and without SMI.
The meta-analysis for diabetes did not detect any heterogeneity between inpatient and community samples with schizophreniform illnesses, suggesting consistency between settings despite the sampling bias inherent to inpatient samples. However the result for people with
bipolar disorder was significantly different from that for schizophrenia. The marked heterogeneity score from the hypertension meta-analysis suggested that there was considerable variation between studies which may have arisen from the differing sampling methods and/or
different definitions of hypertension employed in different papers.
Based on our second level of evidence, (namely studies utilizing general population figures for comparison data) the estimated excess risk of diabetes varied between a zero and a fivefold risk . Only three of these studies permitted calculation of confidence intervals for diabetes
risk ratios and two of these were not significant at the 5% level. This wide variation in the magnitude of diabetes risk between studies may reflect differences in 1) the sampling and definition of SMI, 2) the source of the comparison group (and their inherent risk for diabetes),
and 3) the definition of diabetes or hyperglycemic outcomes. Furthermore, studies that rely on identification of cardiovascular risk factors in routine clinical practice may be flawed due to differential screening rates in people with and without SMI. In the past, people with SMI may have been less likely to be screened for diabetes. More recently there is evidence that people prescribed certain second generation antipsychotics are more likely to receive screening for diabetes. The direction of bias due to differential screening rates may therefore extend in either direction, leading to underestimation or overestimation of diabetes prevalence in SMI, compared to people without.
This review included several small studies that may have lacked statistical power to detect real differences in risk factors or the metabolic syndrome. Few studies have investigated effect modification by age, but there is some support of this phenomenon when comparing people
with and without SMI [3,13].
Strengths and limitations
This is the first review to systematically appraise quality and synthesise data from comparative studies of diabetes, hypertension and lipid levels in people with and without SMI.
We paid critical attention to the quality of studies, in terms of the representativeness of samples, the outcomes measured and we present a large volume of comparative results regarding the prevalence of cardiovascular risk in SMI.
We have grouped these studies according to levels of quality within additional file 2; tables 1–5, especially regarding their selection of comparison data and where possible explored the role of different diagnoses and sampling methods in the meta-analysis. There were insufficient papers to allow us to further subdivide the results.
The review was labour intensive, and like other systematic reviews there was inevitably a delay between the search and publication of the review. The search strategy retrieved over 14,000 papers. However narrowing the search terms was not acceptable because the restricted
search missed several important papers of which we were aware.
Therefore, during the production of this review further evidence may have emerged subsequent to our original search. From our knowledge of the field we are aware of one quality paper meeting our criteria involving both SMI and controls which was published in 2007. Mackin et al
[49] reported data consistent with the main findings of this review. They compared metabolic parameters in 90 people with severe mental illnesses and 92 without. They report increased rates of cardiovascular risk factors in SMI including impaired glucose metabolism, lower HDL cholesterol and raised LDL cholesterol, raised triglycerides and increased metabolic syndrome. In common with our findings, blood pressure was not raised in this study [49].
These recent results have not been included in our analysis because this would bias the systematic nature of our review. Legitimate inclusion of this paper would require re-running of the search for other papers from 2007 and reviewing potentially thousands of new titles. This would be beyond the scope of our current funding.
We acknowledge the difficulty of synthesizing data from multiple studies. In this field many existing studies of cardiovascular risk factors have been opportunistic and have not ensured their SMI samples are representative nor that comparison data are comparable in terms of ethnicity and socio-economic deprivation. This may further explain the observed variation in results and heterogeneity. Furthermore several papers which are commonly cited as evidence
for increased cardiovascular risk in SMI could not be included as they contained insufficient data or no comparison data.
In particular, studies that rely on clinical diagnoses for outcome definition are problematic, since many people may have cardiovascular risk which is undetected (such as abnormal lipids). Screening for lipids, blood pressure and glucose probably occurs in less than a third of people with SMI during routine practice [50]. These points may explain why other narrative reviews conclude the risk of diabetes may be even higher in people with SMI [7,8].
A further challenge is the employment of different definitions of outcomes such as diabetes, hypertension and metabolic syndrome in different studies. We minimised this problem by only including papers which compare risks in people with and without SMI using the same definition,
thus focusing on the relative rather than absolute risk. The relative risk is less sensitive to the employment of differing definitions.
The poor epidemiological quality of studies in this field has been highlighted by a complementary systematic review [51] examining the relative diabetogenic risk of first and second generation antipsychotics. The authors found methodological weaknesses in most studies and were only able to make tentative conclusions about the possible role of second generation antipsychotics in the aetiology of diabetes.
Explanation of excess risk
No studies longitudinally assessed predictors of diabetes (or other cardiovascular risk factors) in SMI. The relative contribution of second generation antipsychotics [2,4,6], lifestyle[10], family history [10], social deprivation and SMI itself (perhaps through chronic stress models) are still
debated [7]. This possible role of SMI itself is supported by two small studies suggesting metabolic disturbances may be observable in newly diagnosed or drug naïve people with SMI [52,53]. However our findings reveal that we do not have an accurate estimate of the contribution of either second generation antipsychotics or SMI itself, to support or refute these theories. We know that the relative risk of metabolic harm differs between the second generation antipsychotics [2,4,6,10] but the absolute attributable risks are unclear.
To understand excess cardiovascular disease in SMI we must improve our knowledge of cardiovascular risk factors in two areas. First, we must accurately determine the extent of the excess of diabetes, dyslipidaemia, hypertension and the metabolic syndrome in representative samples of people with and without SMI, taking into account the effects of age, gender and socio-economic status. This can be achieved in studies with adequate comparison data that explore the level of risk apportionable to SMI and to relevant confounders, especially socio-economic status
[13]. Secondly, to determine the risk attributable to antipsychotic medication, lifestyle, stress and addictions to cardiovascular risk we require data from carefully designed prospective studies. Retrospective estimates of such exposures in SMI are limited by inaccurate reporting and recall bias. Finally, these studies should be adequately powered and should test specific mechanistic hypotheses, rather than measuring multiple risk factors.

Conclusion
Our findings emphasise the importance of poor physical health outcomes in people with SMI, including adverse cardiovascular outcomes. However, we have highlighted gaps in our current knowledge base. This review suggests that diabetes is indeed more common in SMI. Metabolic
syndrome may also be more common, while there is far weaker evidence regarding dyslipidaemia and hypertension. We require high quality studies in representative
samples of people with SMI in which all participants have been screened for cardiovascular risk. These should be cross-referenced to contemporary comparison data regarding the incidence of risk factors in the general population. Furthermore studies should explore differences in
risk factors at different ages and require the statistical power to do so.
The mechanism underlying adverse cardiovascular outcomes remains poorly understood and it is premature to quantify the roles of antipsychotic medication, social adversity, psychiatric symptoms, physiological stress, smoking and diet on the causal pathway of cardiovascular
diseases.

Abbreviations
CHD: Coronary Heart Disease; HDL: High density lipoprotein
(cholesterol); LDL: Low density lipoprotein (cholesterol);
RR: Relative Risk; SMD: Standardised mean
difference; SMI: Severe mental illness.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
DO, IN and MK had the original idea for the study. The
protocol was designed and refined by them along with
CW and GL. The papers and data for the review were
retrieved, reviewed and analysed by all authors including
RD. GL led the analysis DO and CW wrote the initial draft
and tables. All authors commented substantially on subsequent
versions of the manuscript and all authors have
approved the final version.
Funding
The study was funded by a Trial Platform grant from the
UK Medical Research Council. Reference: G0301032

Acknowledgements
Rosalind Lai, Librarian, gave invaluable advice and assistance with the design
and execution of the search strategy

References
1. Lawrence DM, Holman CD, Jablensky AV, Hobbs MS: Death rate
from ischaemic heart disease in Western Australian psychiatric
patients 1980–1998. British Journal of Psychiatry 2003,
182:31-6.
2. Hennekens CH, Hennekens AR, Hollar D, Casey DE: Schizophrenia
and increased risks of cardiovascular disease. Am Heart J
2005, 150:TT1115-1121.
3. Osborn DPJ, Levy G, Nazareth I, Petersen I, Islam A, King M: Relative
risk of cardiovascular and cancer mortality in people
with severe mental illness from the UK General Practice
Research Database. Archives of General Psychiatry 64:242-249.
4. de Leon , Diaz : A meta-analysis of worldwide studies demonstrates
an association between schizophrenia and tobacco
smoking behaviors. Schizophrenia Research 2005, 76:135-157.
5. Marder SR, Essock SM, Miller AL, Buchanan RW, Casey DE, Davis JM,
Kane JM, Lieberman JA, Schooler NR, Covell N, Stroup S, Weissman
EM, Wirshing DA, Hall CS, Pogach L, Pi-Sunyer X, Bigger JT Jr, Friedman
A, Kleinberg D, Yevich SJ, Davis B, Shon S: Physical health
monitoring of patients with schizophrenia. American Journal of
Psychiatry 2004, 161:1334-1349.
6. Anonymous: Which atypical antipsychotic for schizophrenia?
Drug and Therapeutics Bulletin 2004, 42(8):57-62.
7. Thakore JH: Metabolic syndrome and schizophrenia. British
Journal of Psychiatry 2005, 186:435-436.
8. Dinan TG: Schizophrenia and diabetes 2003: an expert consensus
meeting. Introduction. British Journal of Psychiatry 2004,
184(suppl 47):s53-54.
9. McCreadie R, The Scottish Schizophrenia lifestyle group: Diet,
smoking and cardiovascular risk in people with schizophrenia.
British Journal of Psychiatry 2003, 183:534-539.
10. Holt RIG, Bushe C, Citrome L: Diabetes and schizophrenia 2005:
are we any closer to understanding the link? J Psychopharmacol
2005, 19:56-65.
11. StataCorp: Stata Statistical Software: release 8.2. College Station,
TX: Stata Corporation; 2003.
12. Sokal J, Messias E, Dickerson FB, Kreyenbuhl J, Brown CH, Goldberg
RW, Dixon LB: Comorbidity of medical illnesses among adults
with serious mental illness who are receiving community
psychiatric services. Journal of Nervous and Mental Disease 2004,
192(6):421-427.
13. Osborn DPJ, Nazareth I, King MB: Risk for coronary heart disease
in people with severe mental illness: Cross-sectionalcomparative study in primary care. Br J Psychiatry 2006,
188:271-277.
14. Curkendall SM, Mo J, Glasser DB, Stang MR, Jones JK: Cardiovascular
disease in patients with schizophrenia in Saskatchewan,
Canada. J Clin Psychiatry 2004, 65(5):715-720.
15. Enger C, Weatherby L, Reynolds RF, Glasser DB, Walker AM: Serious
cardiovascular events and mortality among patients
with schizophrenia. Journal of Nervous and Mental Disease 2004,
192(1):19-27.
16. Kilbourne AM, Cornelius JR, Han X, Pincus HA, Shad M, Salloum I,
Conigliaro J, Haas GL: Burden of general medical conditions
among individuals with bipolar disorder. Bipolar Disorders 2004,
6:368-373.
17. Finney GOH: Juvenile onset diabetes and schizophrenia? The
Lancet 1989, 2(8673):1214-1215.
18. Makikyro T, Karvonen JT, Hakko H, Nieminen P, Joukamaa M, Isohanni
M, Jones P, Jarvelin M-R: Comorbidity of hospital-treated
psychiatric and physical disorders with special reference to
schizophrenia: A 28 year follow-up of the 1966 Northern Finland
general population cohort. Public Health 1998,
112:221-228.
19. Ryan MCM, Collins P, Thakore JH: Impaired fasting glucose tolerance
in first-episode, drug-naïve patients with schizophrenia.
Am J Psychiatry 2003, 160(2):284-289.
20. Saari KM, Lindeman SM, Viilo KM, Isohanni MK, Jarvelin M-R, Lauren
LH, Savolainen MJ, Koponen HJ: A 4-fold risk of metabolic syndrome
in patients with schizophrenia: The Northern Finland
1966 Birth Cohort Study. J Clin Psychiatry 2005, 66(5):559-563.
21. Arranz B, Rosel P, Ramirez N, Duenas R, Fernandez P, Sanchez JM,
Navarro MA, San L: Insulin resistance and increased leptin concentrations
in noncompliant schizophrenia patients but not
in antipsychotic-naïve first-episode schizophrenia patients. J
Clin Psychiatry 2004, 65(10):1335-1342.
22. Lamberti JS, Crilly JF, Maharaj K, Olson D, Wiener K, Dvorin S,
Costea GO, Bushey MP, Dietz MB: Prevalence of diabetes mellitus
among outpatients with severe mental disorders receiving
atypical antipsychotic drugs. J Clin Psychiatry 2004,
65(5):702-706.
23. Dickerson FB, Pater A, Origoni AE: Health behaviours and health
status of older women with schizophrenia. Psychiatric Services
2002, 53(7):882-884.
24. Casadebaig F, Philippe A, Guillaud-Bataille JM, Gausset MF, Quemada
N, Terra JL: Schizophrenic patients: Physical health and access
to somatic care. Eur Psychiatry 1997, 12:289-293.
25. Dixon L, Weiden P, Delahanty J, Goldberg R, Postrado L, Lucksted A,
Lehman A: Prevalence and correlates of diabetes in national
schizophrenia samples. Schizophrenia Bulletin 2000,
26(4):903-912.
26. Lilliker SL: Prevalence of diabetes in a manic-depressive population.
Comprehensive Psychiatry 1980, 21(4):270-275.
27. Regenold WT, Thapar RK, Marano C, Gavirneni S, Kondapavuluru
PV: Increased prevalence of type 2 diabetes mellitus among
psychiatric inpatients with bipolar affective and schizoaffective
disorders independent of psychotropic drug use. Journal
of Affective Disorders 2002, 70:19-26.
28. Hung C-F, Wu C-K, Lin P-Y: Diabetes mellitus in patients with
schizophrenia in Taiwan. Progress in Neuro-Psychopharmacology &
Biological Psychiatry 2005, 29:523-527.
29. Cohen D, Puite B, Dekker J, Gispen de Wied C: Diabetes mellitus
in 93 chronic schizophrenic inpatients. Eur J Psychiatry 2003,
17(1):38-47.
30. Cassidy F, Ahearn E, Carroll BJ: Elevated frequency of diabetes
mellitus in hospitalized manic-depressive patients. Am J Psychiatry
1999, 156(9):1417-1420.
31. Mukherjee S, Decina P, Bocola V, Saraceni F, Scapicchio PL: Diabetes
mellitus in schizophrenic patients. Comprehensive Psychiatry
1996, 37(1):68-73.
32. Subramaniam M, Chong S-A, Pek E: Diabetes mellitus and
impaired glucose tolerance in patients with schizophrenia.
Can J Psychiatry 2003, 48(5):345-347.
33. Chafetz L, White MC, Collins-Bride G, Nickens J: The poor general
health of the severely mentally ill: Impact of schizophrenic
diagnosis. Community Mental Health Journal 2005, 41(2):169-184.
34. Gierz M, Jeste DV: Physical comorbidity in elderly veterans
affairs patients with schizophrenia and depression. Am J Ger
Psychiatry 1993, 1(2):165-170.
35. Kessing LV, Nilsson FM, Siersma V, Anderson PK: Increased risk of
developing diabetes in depressive and bipolar disorders? J
Psychiatric Research 2004, 38:395-402.
36. Steinert T, Wolfersdorf M, Thoma H, Marpert M: Beeinflusst Dauerhospitalisierung
die kardiovaskuläre Morbidität bei Schizophrenen?
[Does long-term hospitalisation influence the
cardiovascular morbidity in schizophrenics?]. Fortschr neurol
Psychiat 1996, 64:212-220.
37. Cohn T, Prud'homme D, Streiner D, Kameh H, Remington G: Characterizing
coronary heart disease risk in chronic schizophrenia:
High prevalence of the metabolic syndrome. Can J
Psychiatry 2004, 49(11):753-760.
38. Yates WR, Wallace R: Cardiovascular risk factors in affective
disorder. J Affect Disord 1987, 12(2):129-134.
39. Saari K, Jokelainen J, Veijola J, Koponen H, Jones PB, Savolainen M,
Jarvelin MR, Lauren L, Isohanni M, Lindeman S: Serum lipids in
schizophrenia and other functional psychoses: A general
population northern Finland 1966 birth cohort survey. Acta
Psychiatr Scand 2004, 110:279-285.
40. Scottish Schizophrenia Research Group, McCreadie RG, Paterson JR,
Blacklock C, Wiles D, Hall DJ, Graham J, McDonald S, Morrison G,
Mitchell A, Allardyce J, McKane J, Hughson M, Kelly C, Connoly M,
Turner M, Patience D, Yousef A, Brown K, McCallum S, Hay A:
Smoking habits and plasma lipid peroxide and vitamin E levels
in never-treated first-episode patients with schizophrenia.
Br J Psychiatry 2000, 176:290-293.
41. Roccatagliata G, Balestreri R, Ivaldi M, Arcuri T, Caja A: Alterazioni
dell'assetto plasmatico delle componenti lipidiche e proteiche
nell'esito difettuale schizofrenico [Altered plasma
lipid and protein patterns in schizophrenic patients]. Riv Pat
Nerv Ment 1980, 101:94-105.
42. Heiskanen T, Niskanen L, Lyytikainen R, Saarinen PI, Hintikka J: Metabolic
syndrome in patients with schizophrenia. J Clin Psychiatry
2003, 64(5):575-579.
43. Basu R, Brar JS, Chengappa KNR, John V, Parepally H, Gershon S,
Schlicht P, Kupfer DJ: The prevalence of the metabolic syndrome
in patients with schizoaffective disorder – bipolar
subtype. Bipolar Disorder 2004, 6:314-318.
44. Luty J, Kelly C, McCreadie RG: Smoking habits, body mass index
and risk of heart disease: Prospective 2 1/2 year follow-up of
first episode schizophrenic patients. Journal of Substance Use
2002, 7:15-18.
45. McCreadie RG: Diet, smoking and cardiovascular risk in people
with schizophrenia: Descriptive study. Br J Psychiatry 2003,
183:534-539.
46. Susce MT, Villanueva N, Diaz FD, de Leon J: Obesity and Associated
Complications in Patients with Severe Mental Illness: A
Cross-Sectional Survey. J Clin Psychiatry 2005, 66:167-173.
47. McEvoy JP, Meyer JM, Goff DC, Nasrallah HA, Davis SM, Sullivan L,
Meltzer HY, Hsiao J, Stroup TS, Lieberman JA: Prevalence of the
metabolic syndrome in patients with schizophrenia: Baseline
results from the Clinic Antipsychotics Trials of Intervention
Effectiveness (CATIE) schizophrenia trial and comparison
with national estimates from NHANES III. Schizophrenia
Research 2005, 80:19-32.
48. Hingorani A, Vallance P: (1999) A computer programme for
guiding management of CV risk factors and prescribing. British
Medical Journal 1998, 318:101-5.
49. Mackin P, Bishop DR, Watkinson HM, Gallagher P, Ferrier IN: Metabolic
disease and cardiovascular risk in people treated with
antipsychotics in the community. British Journal of Psychiatry
2007, 191(1):23-29.
50. Barnes TRE, Paton C, Cavanagh M-R, Hancock E, Taylor DM, on
behalf of the UK Prescribing Observatory for Mental Health: A UK
audit of screening for the metabolic side effects of antipsychotics
in community patients. Schizophrenia Bulletin 2007,
33:1397-1401.
51. Smith M, Hopkins D, Peveler RC, Holt RIG, Woodward M, Ismail K:
First- v. second-generation antipsychotics and risk for diabetes
in schizophrenia: systematic review and meta-analysis.
The British Journal of Psychiatry 2008, 192:406-411.

Case 31-2006: A 15-Year-Old Girl with Severe Obesity

Alison G. Hoppin, M.D., Eliot S. Katz, M.D., Lee M. Kaplan, M.D., Ph.D.,
and Gregory Y. Lauwers, M.D.

Presentation of Case
A 15 1/2 -year-old girl was seen in the outpatient Weight Center of this hospital for
the evaluation of severe obesity. She had had a normal gestation without complications
and had been adopted during the first month of life. She weighed 3.9 kg at birth and 4.8 kg at 1 month of age. At the age of 1 year, her weight-to-length ratio was in the 75th percentile. At 3 years of age, her body-mass index (BMI, the weight in kilograms divided by the square of the height in meters) was above the 97th percentile. She was referred to a nutritionist. Her appetite remained steady, and she ate most foods. Although her food intake appeared to be similar to that of the other children in the family, her BMI continued to increase.
Snoring and restless sleep began at approximately 5 to 6 years of age. At the age of 7 years, she was enrolled in a monthly weight-control program, and a year later, she was evaluated by a nutrition specialist. Physical examination revealed an overweight child with mild acanthosis nigricans of the neck with no other abnormalities. The next year, her parents noted that she was eating secretly; hyperpigmentation of the thighs was noted on physical examination. Between the ages of 10 and 11 years, her weight increased approximately 15 kg, and she began a program for weight control at her pediatrician’s practice. At 12 years of age, she was seen by a psychiatrist, who noted dysthymia and poor motivation and prescribed sertraline and psychotherapy.
When the patient was 13 years old, her parents noticed nocturnal somnambulation and an increased intake of food. At 13 years 3 months, the patient was referred to an endocrinologist because she had not lost weight and had chronic daytime fatigue. Her height was 162.4 cm, her weight 106.9 kg, her blood pressure 132/73 mm Hg, and her pulse 76 beats per minute. Breast development was Tanner stage 3, and pubertal development was Tanner stage 5 (with 1 representing immature development and 5 maturity); acanthosis nigricans was present around
the neck and groin. The remainder of the examination was normal; there was no hirsutism. Snoring and insomnia worsened during adolescence. Menarche occurred at the age of 14 years, and her menstrual cycles were irregular. Daytime sleepinessworsened, including falling asleep at school, and morning headaches developed. A combination of dextroamphetamine sulfate and amphetamine aspartate was prescribed to enhance alertness. During an evaluation by a sleep specialist at the age of 14 years, physical examination revealed a blood pressure of 120/90 mm Hg; the tonsils were enlarged, but there was no marked crowding. A series of overnight polysomnograms obtained between the ages of 13 and 15 years showed progressive
worsening of obstructive sleep apnea, including intermittent oxygen desaturation, hypercapnia,
periodic leg movements, and sleep disruption. Bilevel ventilation therapy was started when the
patient was 13 years old, but compliance with the therapy was poor. Pulmonary-function testing revealed normal spirometric values and lung volumes. The results of electrocardiography, chest
radiography, and echocardiography were normal.

At the age of 14 years (17 months before this evaluation), the patient attended a summer camp and lost approximately 20 kg; she promptly regained the weight after returning home and gained an additional 12 kg during the subsequent year. Three months before presentation, she participated in a home-administered weight-loss planon the basis of a point system, but her weight continued to increase.

At the time of the evaluation in the Weight Center, she had daytime somnolence but no headaches. She drank low-calorie soft drinks and two glasses of juice daily. She snacked twice during the night on sandwiches or other carbohydratecontaining foods. She exercised with a personal trainer three to five times per week and watchedShe typically slept 7 hours on school nights and 12 hours per night on the weekends. She had no difficulty initiating sleep, but she was hard to arouse in the morning. The patient’s early development had been normal; she walked at 12 months of age and spoke in short sentences at 15 months. Her depression had improved after a change in schools during the year preceding her presentation at the Weight Center, and the psychotherapy was discontinued.

She was a good student in the 10th grade. A grandmother in her birth family was known to
have been overweight; no other biologic-family history was known. Members of her adoptive family, including her parents and two younger siblings, were of normal weight. Her only medication was sertraline, and she had no known allergies.
Her height was 164.5 cm, her weight 126.6 kg, and her BMI 46.7. The blood pressure was 124/95 mm Hg. Severe acanthosis nigricans was present on the neck and axillae, and there was moderate acne on the face; a slightly android pattern of hair growth was evident on the abdomen, and there were moderate striae on the lower abdomen. There was no hair growth on the face, no rash in the skin folds, and no edema. The remainder of the physical examination was normal.


Differential Diagnosis
Dr. Alison G. Hoppin: This patient presented with uncommon manifestations of a common disease. Obesity is common: 17.4% of adolescents in the United States are considered overweight by the standards of the Centers for Disease Control and Prevention.1 However, this patient’s degree of obesity was very unusual: with a BMI of 46.7, she had adiposity levels that constituted class 3 obesity (on a scale of 1 to 3, with class 1 indicating a BMI of 30.0 to 34.9, class 2 a BMI of 35.0 to 39.9, and class 3 a BMI of 40.0 or more) in an adult. In addition, she had most of the important medical complications of obesity in children and adolescents.
This patient’s most acute health issues at presentation were symptoms suggestive of sleep apnea and diabetes mellitus. Because the symptoms of sleep apnea are somewhat subjective and there are no clear screening criteria, the problem is probably underdiagnosed in children and adolescents with obesity.2-5 This patient was referred to Dr. Eliot Katz, a specialist in sleep disorders in children, who will discuss the evaluation and management of her sleep apnea.
Obstructive Sleep Apnea

Dr. Eliot S. Katz: Testing of this patient by overnight polysomnography at 15 years of age
indicated that her sleep latency was less than4 minutes (normal, 9 to 33), suggesting objective
sleepiness. She had recurrent episodes of partial or complete upper-airway obstruction associated with intermittent hypoxemia (minimum oxygen saturation, 86%; normal value, 92 to 96), hypercapnia (awake, 52 mm Hg, and asleep, 64 mm Hg; normal carbon dioxide peak during sleep, ≤53 mm Hg), and electroencephalographic arousal . Her apnea–hypopnea index was markedly elevated at 21 events per hour (normal value, ≤1). Despite these findings, she had normal sleep architecture and sleep efficiency. Children with severe obstructive sleep apnea often have normal distribution of sleep states, despite frequent episodes of obstruction and brief electrocortical arousal.
Obesity poses both an obstructive load to the upper airway and an elastic load to the entire pulmonary system. Although pulmonary function, as measured by spirometry, is often normal in obese children during wakefulness at rest, as it was in this child, there are often measurable deficits during exercise and sleep. Obese children are 4.5 times as likely to have obstructive sleep apnea as are children who are not obese.[19] The severity of the condition is related to the degree of visceral adiposity, rather than to the amount of total body fat. This patient had a central pattern of obesity, which has the strongest correlation with the metabolic syndrome.[20] More than 90% of children with both obesity and habitual snoring have obstructive
sleep apnea.21 Thus, this patient probably had obstructive sleep apnea during her 8-to-10-
year history of snoring before her initial polysomnography. Androgens affect ventilatory control
and increase visceral fat; thus, obstructive sleep apnea is more commonly seen in boys after puberty (rather than before puberty) and in women who have excessive levels of androgen associated with the polycystic ovary syndrome, which was suspected in this patient.[22]
Sequelae
The consequences of obstructive sleep apnea include cardiovascular abnormalities, neurocognitive impairment, daytime sleepiness, and metabolic disturbances — several of which were seen in this patient. Her hypertension was probably a consequence of both obstructive sleep apnea and obesity. In extreme cases, biventricular dysfunction and pulmonary hypertension may develop, and screening for pulmonary hypertension withsevere obstructive sleep apnea. The 8-to-10-year history of probable obstructive sleep apnea during a vulnerable period of brain growth places this patient at risk for neurocognitive impairment, including mood disturbance, which she had. The disease has a minimal effect on intelligence but may impair executive functioning and attention span and has been associated with poor school performance in adolescents.[23]
In contrast to adults with sleep apnea, most children do not have clinical or objective sleepiness.
24 Sleepiness in adolescents is frequently multifactorial and includes an insufficient amount
of sleep, poor sleep hygiene, increased sleep requirements and circadian-rhythm disturbances,
and disrupted sleep related to obstructive sleep apnea, sleepwalking, and nocturnal eating. Thus, this patient had many reasons for sleepiness.
Both obstructive sleep apnea and obesity are associated with increased levels of inflammatory
markers,[20,25] which are believed to be involved inthe normal homeostatic regulation of sleep. Thus, increased levels of circulating cytokines may contribute to the sleepiness that is characteristic of both obstructive sleep apnea and obesity. Obstructive sleep apnea can increase insulin resistance and leptin levels, independent of obesity,[20,25] and treatment with continuous positive airway pressure can reduce insulin resistance and levels of leptin and inflammatory markers.
Treatment
In contrast to children of normal weight, obese children with adenotonsillar hypertrophy typically do not have complete resolution of obstructive sleep apnea after adenotonsillectomy, although their condition typically improves.[26] In this patient, nasopharyngoscopy demonstrated only mild adenotonsillar hypertrophy that was not considered to warrant surgical removal. Weight loss can lessen the severity of obstructive sleep apnea, but residual obstruction is frequently present.[27]

Continuous positive airway pressure and bilevel ventilation are both effective therapies for obstructive sleep apnea.28 However, 15 to 20% of patients will not comply with the use of nasal positivepressure ventilation at all; for the remainder of patients, the average duration of use is approximately 5 hours per night. Although this patient acknowledged improvements in daytime functioning after receiving nocturnal bilevel ventilation, she complied poorly with therapy. Thus, her case illustrates both the consequences of sleep apnea and the difficulties in managing it.
Dr. Hoppin: This patient had features of the metabolic syndrome, a constellation of findings
associated with an increased risk of atherosclerotic cardiovascular disease and type 2 diabetes
mellitus.[10]
Insulin Resistance and Type 2 Diabetes Mellitus
This patient had evidence of insulin resistance, with severe acanthosis nigricans, an elevated insulin level after an overnight fast, and a glycated hemoglobin value of 6.7%. Subsequent testingrevealed a blood glucose level of 141 mg per deciliter (7.8 mmol per liter) 2 hours after an oral glucose load , documenting impaired glucose tolerance.[29] Insulin resistance is independently associated with both obesity and puberty, so this patient is at risk for both reasons.[6-8] In a longitudinal study of obese adolescents, progression from normal to impaired glucose tolerance occurred in 10% during a 2-year period, and 25% of these patients had progression to type 2 diabetes.[30]
Six months after the first visit, treatment with metformin was begun, in consultation with an
endocrinologist. During the next 2 years, the patient had a slight improvement in glycemic control, with glycated hemoglobin values ranging from 6.5 to 7% and fasting blood glucose levels between 90 and 95 mg per deciliter (5.0 to 5.3 mmol per liter). However, during a period of noncompliance with the metformin, her fasting blood glucose levels rose to a range of 244 to 275 mg per deciliter (13.5 to 15.3 mmol per liter), diagnostic of diabetes mellitus.
Hypertension and Dyslipidemia
The patient had mild hypertension at her first visit, which gradually worsened during the next
18 months. Pharmacologic intervention is recommended for hypertension that persists despite
a modification in diet and for patients with diabetes mellitus.[9] Angiotensin-converting–enzyme inhibitors are recommended preferentially in children with diabetes. Treatment with lisinopril
was initiated 18 months after her first visit. Her fasting total cholesterol and triglyceride levels
were high, whereas the level of low-density lipoprotein cholesterol was in the borderline range.
Our dietary counseling included recommendations for a reduction in dietary fat.[31]
Polycystic Ovary Syndrome
The patient also had findings that suggested the polycystic ovary syndrome and nonalcoholic fatty liver disease, both of which are associated with the metabolic syndrome phenotype. The classic clinical features of the polycystic ovary syndrome include menstrual disturbance, hirsutism, and polycystic ovaries. However, on the basis of broader diagnostic criteria, the disorder is thought to affect 5 to 10% of women of reproductive age.[11] Although the patient had menarche and menstrual patterns that could be considered normal and shedid not have marked hirsutism, she had an elevated level of free testosterone at 13 years of age, suggesting the presence of hyperandrogenism. Two years after her first visit here, she presented with acute right ovarian torsion due to a cyst. During surgery to remove the cyst, the contralateral
ovary was polycystic on gross examination, confirming the diagnosis of the polycystic ovary syndrome. After the operation, treatment with ethinyl estradiol and drospirenone was begun.
About 30% of adolescent girls with the polycystic ovary syndrome have glucose intolerance
or diabetes mellitus.[32] Hyperinsulinemia seems to be the common mechanism: insulin acts synergistically with luteinizing hormone to increase the production of androgen by the ovarian theca cells while also decreasing the level of sex hormone– binding globulin.[12,33] Treatment with metformin can lead to clinical improvement, even in adolescents without overt diabetes mellitus. [34]
Nonalcoholic Fatty Liver Disease
On initial evaluation, mild elevations of serum aminotransferase levels were present, without hyperbilirubinemia. Nonalcoholic fatty liver disease is the most common cause of mild aminotransferase elevations in children and is strongly associated with obesity and the metabolic syndrome.[13,14] It is important to exclude other causes of liver disease, so when the finding persisted, we did laboratory testing to rule out viral hepatitis, autoimmune hepatitis, and Wilson’s disease; all the test results were negative. No specific treatments have been established for fatty liver disease in children or adults, but weight loss is almost certainly helpful.
Causes of Obesity
We have discussed the consequences of the patient’s obesity, but what can be said about the
causes of it? Because she was adopted in early infancy, her case illustrates better than most how
biologic determinants of obesity can dominate over lifestyle or environmental exposures, as Dr.
Kaplan will discuss.
Dr. Lee M. Kaplan: In obesity, a combination of genetic, developmental, and environmental determinants alters the body’s normal system for the regulation of weight. The prevalence of obesity has increased in the past 50 years, with a disproportionate increase in severe obesity. Between 1986 and 2000, the prevalence of obesity (BMI >30) increased by a factor of 2, the prevalence of class 3 obesity (BMI >40) increased by a factor of 4, and the prevalence of the most severe forms of obesity (BMI >50) increased by a factor of 6.[35,36]
Most people are genetically susceptible to abnormal weight gain under the right conditions.
The availability of highly processed, calorie-dense foods and a decreasing level of physical activity are important environmental contributors to obesity, but they are not the only ones. Disrupted meal patterns, disordered and inadequate sleep, disturbances in normal circadian rhythms, high levels of stress, social isolation, and the use of medications that promote weight gain may be equally important. Several of these factors may have affected the patient, including sleep deprivation and the stresses of adolescence. In contrast, the contribution of her diet or pattern of eating to the obesity appears to be limited.
The early onset of obesity and the striking difference in weight between the patient and her
adoptive siblings despite similar eating habits suggest an important contribution of genetics to her obesity.37,38 There is strong evidence that genetic background plays an important role in determining the predisposition to obesity. Obesity often runs in families, and the correlation of BMI among siblings is little affected by whether they were raised together or apart and is often independent of the type or pattern of food intake or physical activity. Studies of twins suggest that genetic factors determine about 50 to 70% of the predisposition to the development of obesity.[39]
We have little information about the biologic relatives of the patient, other than that her grandmother was overweight. Although the onset of her obesity was very early, most of the known monogenetic or oligogenetic causes of obesity are unlikely. [40] These rare disorders reflect alterations in genes that encode central nervous system regulators of body weight, such as leptin, the leptin receptor, melanocyte-stimulating hormone, and the melanocortin 4 receptor (MC4R). Genetic testing through a research protocol revealed no evidence of an abnormal MC4R in the patient. The plasma leptin level was 47 ng per milliliter (normal range, 3.3 to 18.3); the elevated level was appropriate to her obesity, thereby excluding genetic deficiency in leptin or its receptor. Symptoms of the most common, well-defined obesity syndromes — including the Prader–Willi syndrome, the Bar-det–Biedl syndrome, and a deficiency in singleminded
homologue 1 (SIM-1) — were absent.
Many genes contribute to the regulation of body weight, and a genetic predisposition to obesity,
which the patient and other children with early-onset obesity almost certainly have, probably
results from the influence of multiple genes, which combine to support an energy-thrifty phenotype. [41]

Discussion of Management
Weight-Loss Strategies
Dr. Hoppin: Despite the likelihood of a strong biologic basis for the patient’s obesity, there is no
specific physiological target that we can address to facilitate her weight management. The patient had lost approximately 20 kg in weight while attending a summer weight-loss camp 17 months before presentation to the Weight Center; she rapidly regained this weight after returning home. Such rebounds after acute weight loss from dietary restriction are very common and probably speak to the resilience of weight-regulatory mechanisms rather than to bad habits or a failure of willpower. Thus, we believed that the lifestyle habits of the patient and family were not the primary cause of her obesity. Nonetheless, our first approach was to work with her and her family to optimize these habits.
This patient embarked on a new series of weight-control attempts, with limited success. She continued to work with a personal trainer, engaging in aerobic and strength training 3 days a week, and received dietary supervision from the trainer, supplemented by individual nutritional
counseling from a registered dietitian. She was able to stabilize her eating patterns to some degree, and the frequency of nocturnal eating decreased. After her first visit, sertraline was
tapered and then discontinued, and a trial of sibutramine was begun 2 months later. Metformin
was added 4 months later for glycemic control. With these combined therapies, her weight
gain stopped, but she lost only 3.2 kg during the first year, and sibutramine was ultimately with drawn.
When she was 17 years old, she participated in a group-based program at our center with
weekly meetings for 3 months. The group consisted of adolescent girls with obesity and was
led by a dietitian. Specific goals were to improvefood choices and planning and included both
nutritional education and behavioral techniques. Dietary guidelines included modest caloric restriction and a relatively low intake of simple carbohydrates. The parents of the girls met concurrently with a psychologist to address family dynamics related to weight control. The patient participated actively and enthusiastically in the group program; she reduced her weight by another 3.2 kg and maintained the weight loss during the subsequent 6 months.
Three years after her first evaluation, the patient had maintained a 6.8-kg weight loss for
about a year. However, her BMI of 44 remained inthe range of severe obesity, and she continued to have sleep apnea, diabetes mellitus, hypertension, and dyslipidemia.
Bariatric Surgery
When the patient was 18 years old, we began to consider the possibility of weight-reduction surgery. Gastric bypass is clearly the most consistently effective treatment for severe obesity in
adults[42,43]; although less is known about outcomes in adolescents, a few published case series
suggest that they are similar to those in adults.[44] An expert panel[45] has recommended that weight reduction surgery be considered for adolescents with a BMI of more than 40 who have severe medical complications of obesity (such as sleep apnea or diabetes, as in this patient) or with a BMI of more than 50 who have any medical complications and in whom efforts to control weight through other measures have been unsuccessful. In adolescents, medical follow-up to monitor for and treat potential micronutrient deficiencies is essential.[46]
The patient was an appropriate candidate for bariatric surgery. Her weight and coexisting illnesses were appropriate indications according to criteria for both adults and adolescents. She had made a variety of sustained efforts to control her weight by other measures, and she had an excellent record of compliance with long-term medical follow-up.
At the age of 19 years, the patient underwent a laparoscopic Roux-en-Y gastric bypass. Her initial postoperative recovery was uncomplicated. Awedge biopsy of the liver was performed during the operation.
Dr. Gregory Y. Lauwers: The liver biopsy showed histologic evidence of nonalcoholic steatohepatitis. The condition is characterized by inflammation and fibrosis, indicating damage to hepatocytes, and is considered to be a progressive form of nonalcoholic fatty liver disease. It has
the potential to progress to cirrhosis and hepatocellular carcinoma, but the magnitude of the
risk is not known.[47]
Dr. Lynne L. Levitsky (Pediatric Endocrinology): At her most recent follow-up, 1 month after surgery, the patient’s weight had decreased from 122.5 kg to 109.6 kg, with a BMI of 39.8. Metformin and lisinopril had been discontinued at the time of the surgery. The blood pressure was 138/79 mm Hg, and her morning blood glucose levels by finger-stick measurement had been normal on all but two occasions. On physical examination, acanthosis nigricans that had developed on the wrists and ankles had resolved but persisted on the neck.

Anatomical Diagnosis
Severe childhood obesity with obstructive sleep apnea, hypertension, impaired glucose tolerance progressing to type 2 diabetes mellitus, the polycystic ovary syndrome, and nonalcoholic steatohepatitis.

Dr. Hoppin reports having received grant support from the
Massachusetts Vitamin Litigation Fund. No other potential conflict
of interest relevant to this article was reported.

References
1.Ogden CL, Carroll MD, Curtin LR, et
al. Prevalence of overweight and obesity
in the United States, 1999-2004. JAMA
2006;295:1549-55.
2.Gislason T, Benediktsdottir B. Snoring,
apneic episodes, and nocturnal hypoxemia
among children 6 months to 6
years old: an epidemiologic study of lower
limit of prevalence. Chest 1995;107:963- 6.
3.Ali NJ, Pitson DJ, Stradling JR. Snoring,
sleep disturbance, and behaviour in
4-5 year olds. Arch Dis Child 1993;68:360-6.
4.Marcus CL, Curtis S, Koerner CB, Joffe
A, Serwint JR, Loughlin GM. Evaluation
of pulmonary function and polysomnography
in obese children and adolescents.
Pediatr Pulmonol 1996;21:176-83.
5.Chay OM, Goh A, Abisheganaden J, et
al. Obstructive sleep apnea syndrome inobese Singapore children. Pediatr Pulmonol
2000;29:284-90.
6.Dolan LM, Bean J, D’Alessio D, et al.
Frequency of abnormal carbohydrate metabolism
and diabetes in a populationbased
screening of adolescents. J Pediatr
2005;146:751-8.
7.Sinha R, Fisch G, Teague B, et al.
Prevalence of impaired glucose tolerance
among children and adolescents with
marked obesity. N Engl J Med 2002;346:
802-10. [Erratum, N Engl J Med 2002;346:
1756.]
8.Williams DE, Cadwell BL, Cheng YJ,
et al. Prevalence of impaired fasting glucose
and its relationship with cardiovascular
disease risk factors in US adolescents,
1999-2000. Pediatrics 2005;116:1122-6.
9.National High Blood Pressure Education
Program Working Group on High
Blood Pressure in Children and Adolescents. The fourth report on the diagnosis,
evaluation, and treatment of high blood
pressure in children and adolescents. Pediatrics
2004;114:Suppl:555-76.
10.Cook S, Weitzman M, Auinger P, Nguyen
M, Dietz WH. Prevalence of a metabolic
syndrome phenotype in adolescents:
findings from the Third National Health
and Nutrition Examination Survey, 1988-
1994. Arch Pediatr Adolesc Med 2003;157:
821-7.
11.Dunaif A. Insulin resistance and the
polycystic ovary syndrome: mechanism
and implications for pathogenesis. Endocr
Rev 1997;18:774-800.
12.Ehrmann DAA. Polycystic ovary syndrome.
N Engl J Med 2005;352:1223-36.
13.Franzese A, Vajro P, Argenziano A, et
al. Liver involvement in obese children:
ultrasonography and liver enzyme levels
at diagnosis and during follow-up in anItalian population. Dig Dis Sci 1997;42:
1428-32.
14.Strauss RS, Barlow SE, Dietz WH. Prevalence
of abnormal serum aminotransferase
values in overweight and obese adolescents.
J Pediatr 2000;136:727-33.
15.Jerre R, Karlsson J, Henrikson B. The
incidence of physiolysis of the hip: a population-
based study of 175 patients. Acta
Orthop Scand 1996;67:53-6.
16.Cook SD, Lavernia CJ, Burke SW,
Skinner HB, Haddad RJ Jr. A biomechanical
analysis of the etiology of tibia vara.
J Pediatr Orthop 1983;3:449-54.
17.von Mutius E, Schwartz J, Neas LM,
Dockery D, Weiss ST. Relation of body
mass index to asthma and atopy in children:
the National Health and Nutrition
Examination Study III. Thorax 2001;56:
835-8.
18.Balcer LJ, Liu GT, Forman S, et al. Idiopathic
intracranial hypertension: relation
of age and obesity in children. Neurology
1999;52:870-2.
19.Redline S, Tishler PV, Schluchter M,
Aylor J, Clark K, Graham G. Risk factors
for sleep-disordered breathing in children:
associations with obesity, race, and respiratory
problems. Am J Respir Crit Care Med
1999;159:1527-32.
20.Vgontzas AN, Papanicolaou DA, Bixler
EO, et al. Sleep apnea and daytime sleepiness
and fatigue: relation to visceral obesity,
insulin resistance, and hypercytokinemia.
J Clin Endocrinol Metab 2000;85:
1151-8.
21.Silvestri JM, Weese-Meyer DE, Bass MT,
Kenny AS, Hauptman SA, Pearsall SM.
Polysomnography in obese children with
a history of sleep-associated breathing
disorders. Pediatr Pulmonol 1993;16:124- 9.
22.Vgontzas AN, Legro RS, Bixler EO,
Grayev A, Kales A, Chrousos GP. Polycystic
ovary syndrome is associated with
obstructive sleep apnea and daytime sleepiness:
role of insulin resistance. J Clin Endocrinol
Metab 2001;86:517-20.
23Gozal D, Pope DW Jr. Snoring during
early childhood and academic performance
at ages thirteen to fourteen years. Pediatrics
2001;107:1394-9.
24.Gozal D, Wang M, Pope DW Jr. Objective
sleepiness measures in pediatric obstructive
sleep apnea. Pediatrics 2001;108:
693-7.
25.Weiss R, Dziura J, Burgert TS, et al.
Obesity and the metabolic syndrome in
children and adolescents. N Engl J Med
2004;350:2362-74.
26.Wiet GJ, Bower C, Seibert R, Griebel
M. Surgical correction of obstructive sleep
apnea in the complicated pediatric patient
documented by polysomnography. Int J Pediatr
Otorhinolaryngol 1997;41:133-13.
27.Dixon JB, Schachter LM, O’Brien PE.
Polysomnography before and after weight
loss in obese patients with severe sleep
apnea. Int J Obes (Lond) 2005;29:1048-54.
28.Marcus CL, Ward SL, Mallory GB, et al.
Use of nasal continuous positive airway
pressure as treatment of childhood obstructive
sleep apnea. J Pediatr 2002;127:88-94.
29.American Diabetes Association. Diagnosis
and classification of diabetes mellitus.
Diabetes Care 2004;27:Suppl 1:S5-S10.
30.Weiss R, Taksali SE, Tamborlane WV,
Burgert TS, Savoye M, Caprio S. Predictors
of changes in glucose tolerance status in
obese youth. Diabetes Care 2005;28:902-9.
31.American Academy of Pediatrics, Committee
on Nutrition. Cholesterol in childhood.
Pediatrics 1998;101:141-7.
32.Palmert MR, Gordon CM, Kartashov
AI, Legro RS, Emans SJ, Dunaif A. Screening
for abnormal glucose tolerance in adolescents
with polycystic ovary syndrome.
J Clin Endocrinol Metab 2002;87:1017-23.
33.Dunaif A. Hyperandrogenic anovulation
(PCOS): a unique disorder of insulin
action associated with an increased risk
of non-insulin-dependent diabetes mellitus.
Am J Med 1995;98:Suppl 1A:33S-39S.
34.Ibanez L, Ferrer A, Ong K, Amin R,
Dunger D, de Zegher F. Insulin sensitization
early after menarche prevents progression
from precocious pubarche to polycystic
ovary syndrome. J Pediatr 2004;144:23-9.
35.Flegal KM, Carroll MD, Ogden CL,
Johnson CL. Prevalence and trends in obe-
sity among US adults, 1999-2000. JAMA
2002;288:1723-7.
36.Ogden CL, Carroll MD, Curtin LR,
McDowell MA, Tabak CJ, Flegal KM. Prevalence
of overweight and obesity in the
United States, 1999-2004. JAMA 2006;295:
1549-55.
37.Bell CG, Walley AJ, Froguel P. The genetics
of human obesity. Nat Rev Genet
2005;6:221-34.
38.Lyon HN, Hirschhorn JN. Genetics of
common forms of obesity: a brief overview.
Am J Clin Nutr 2005;82:Suppl:215S-
217S.
39.Bouchard C, Tremblay A, Després J-P,
et al. The response to long-term overfeeding
in identical twins. N Engl J Med 1990;
322:1477-82.
40.Rosenbaum M, Leibel RL, Hirsch J.
Obesity. N Engl J Med 1997;337:396-407.
[Erratum, N Engl J Med 1998;338:555.]
41.Prentice AM. Early influences on human
energy regulation: thrifty genotypes
and thrifty phenotypes. Physiol Behav
2005;86:640-5.
42.Buchwald H, Avidor Y, Braunwald E,
et al. Bariatric surgery: a systematic review
and meta-analysis. JAMA 2004;292:
1724-37. [Erratum, JAMA 2005;293:1728.]
43.Case Records of the Massachusetts
General Hospital (Case 25-2004). N Engl
J Med 2004;351:696-705.
44.Inge TH, Zeller MH, Lawson ML, Daniels
SR. A critical appraisal of evidence supporting
a bariatric surgical approach to
weight management for adolescents. J Pediatr
2005;147:10-9.
45.Apovian CM, Baker C, Ludwig DS, et al.
Best practice guidelines in pediatric/adolescent
weight loss surgery. Obes Res 2005;
13:274-82.
46.Towbin A, Inge TH, Garcia VF, et al.
Beriberi after gastric bypass surgery in
adolescence. J Pediatr 2004;145:263-7.
47.Kleiner DE, Brunt EM, Van Natta M, et
al. Design and validation of a histological
scoring system for nonalcoholic fatty liver
disease. Hepatology 2005;41:1313-21.