“The Linosa Study”: Epidemiological and heritability data of the metabolic syndrome in a Caucasian genetic isolate

Nutrition, Metabolism & Cardiovascular Diseases (2009) 19, 455e461

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/nmcd

‘‘The Linosa Study’’: Epidemiological and heritability data of the metabolic syndrome in a Caucasian genetic isolate* A. Bellia a, E. Giardina b, D. Lauro a, M. Tesauro a, G. Di Fede c, G. Cusumano c, M. Federici a, G.B. Rini c, G. Novelli b, R. Lauro a, P. Sbraccia a,* a

Department of Internal Medicine, University of Rome ‘‘Tor Vergata’’, Via Montpellier 1, I-00133 Rome, Italy Department of Biopathology and Diagnostic Imaging, University of Rome ‘‘Tor Vergata’’, Italy c Department of Clinical Medicine and Emerging Diseases, University of Palermo, Italy b

Received 17 June 2008; received in revised form 6 October 2008; accepted 10 November 2008

KEY WORDS Metabolic syndrome; Heritability; Obesity; Insulin resistance

Abstract Background and aims: Growing evidence suggests that the metabolic syndrome (MetS) has both a genetic and environmental basis. To evaluate the possibility of a further genetic analysis, we estimated prevalence rates and heritabilities for the MetS and its individual traits in the adult population of Linosa, a small and isolated Italian Island in the southern-central part of the Mediterranean Sea. Methods and results: The Linosa Study (LiS) group consisted of 293 Caucasian native subjects from 51 families (123 parents; 170 offsprings). The MetS was defined according to NCEP/ATP III criteria and the following prevalence rates were calculated: hyperglycaemia 20.3%; central obesity 34.9%; hypertension 43.4%; hypertriglyceridaemia 29.9%; ‘‘low HDL’’ 56.6%; MetS 29.9%. Waist circumference was significantly related to all the quantitative parameters included in the NCEP/ATP III MetS definition. The MetS showed a heritability of 27% (p Z 0.0012) and among its individual components, treated as continuous and discrete traits, heritability ranged from 10% for blood glucose to 54% for HDL-cholesterol. Among MetS subtypes, the clustering of central obesity, hypertriglyceridaemia and ‘‘Iow HDL’’ had the highest heritability (31%; p < 0.001). Conclusion: These data showed high prevalence rates for the MetS and its related traits in an isolated and small Caucasian population. The appreciable heritability estimates for the MetS

* Work supported by grants from the Ministero della Salute (RF 2005econv. N. 76) and from the Ministero dell’Universita ` e della Ricerca (prot. 2005060925_003). This project was also included in the Public Population Project in Genomics (P3G). * Corresponding author. Tel.: þ39 06 72596888; fax: þ39 06 72596890. E-mail address: [email protected] (P. Sbraccia).

0939-4753/$ - see front matter ª 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.numecd.2008.11.002

456

A. Bellia et al. and some of its components/clusters in the LiS population might support the observation of genetic factors underlying the pathogenesis of the MetS and encourage further analysis to identify new susceptibility genes. ª 2008 Elsevier B.V. All rights reserved.

Introduction

Methods

The MetS has been described as a clustering of risk factors associated with coronary heart disease (CHD) and type 2 diabetes [1e3]. The 2004 AHA/NHLBI/ADA conference confirmed CHD as a major clinical outcome of the MetS and identified the six major components: abdominal obesity, atherogenic dyslipidaemia, elevated blood pressure and glucose and a proinflammatory and a prothrombotic state [4,5]. It has been hypothesized that insulin resistance [6,7] and abdominal obesity [8e10] may be the key factors linking the components of the MetS together. Indeed, excessive calorie intake and physical inactivity promote the development of obesity, leading to an impaired metabolic profile, while some individuals are genetically predisposed to insulin resistance [11e14]. Furthermore, the prevalence of the MetS varies among different ethnic groups [15e19]. For example, according to the NHANES III, adult prevalence of the MetS is higher in Hispanic-Americans than in Caucasian-Americans [15]. Similar findings were reported in the Framingham Offspring Study and in the San Antonio Heart Study [16], while a lower prevalence was reported in Caucasian middle-aged men from Finland [17] and Italy [19]. Besides variations in environmental factors and diagnostic criteria, a genetic susceptibility could explain these differences. A powerful approach to mapping the genes for complex diseases, such as the MetS, is to study isolated founder populations, in which extended genealogical data are available and genetic heterogeneity and environmental noise are likely to be reduced compared to out-bred populations [20e22]. To evaluate the possibility of such an analysis we focused on the population of Linosa, a small Italian island closer to the African continent than Sicily, in which the presence of metabolic disorders reported by local physicians had appeared fairly high. Linosa Island was first colonized in 1839 under the Borbonic government, when a first nucleus of people made up of some families from Sicily was sent in order to create a community. The inhabitants have lived on agricultural products and fishing up to the present day and, thanks to marriage among the same members, have reached a population of about 420 people with only 11 family names. Furthermore, Linosa Island has been always reachable only by sea and consistent with the weather conditions. For all these reasons the population of Linosa may be considered a genetically welldefined and isolated Caucasian group, with a high degree of consanguinity and minimal western-style environmental influences. The aims of the present study were: (i) to provide a metabolic profile of the Linosa Island adult population; (ii) to estimate the heritability of the MetS and its individual traits in order to perform a further genetic analysis.

Subjects The baseline evaluation was carried out in Linosa Island between May and July 2005. The overall adult population (>18 years old) of Linosa Island reported by the local registry consisted of 420 subjects. A total of 364 individuals (201 women, 163 men), aged 48  18 years (mean  SD), agreed to participate in the study, corresponding to about 85% of the whole native adult population. All participants were white Southern-Europeans. Among these subjects, 71 individuals with incomplete data collection were excluded from the analysis; thus, the final Linosa Study (LiS) group consisted of 293 adult individuals. For the purpose of this study, using the existing family data and information from the subjects, we reconstructed 51 ‘‘nuclear’’ families, comprising 123 parents (63.2  10.9 years) and 170 offsprings (33.1  9.7). Family sizes ranged from 3 relatives to 10 family members, with an average of 5.2 members per family. All subjects gave their written informed consent to participate in the study. The protocol was approved by the Ethical Committee of the ‘‘Azienda Ospedaliera di Palermo’’.

Clinical and physical examination data The following demographic and clinical data were collected with a standardized questionnaire: sex, age, cigarette smoking, alcohol consumption, physical activity, medical history and medication status. Weight (to the nearest 0.5 kg) and height (to the nearest 0.5 cm) were measured while the subjects were fasting overnight and wearing only underwear. Body mass index (BMI) was calculated as weight (kg) divided by height (m2). Waist circumference (WC) was measured in the standing position with a flexible plastic tape at the level of the iliac crest. Systolic (SBP) and diastolic (DBP) blood pressure were measured in the seating position with a standard mercury sphygmomanometer on the left arm after at least 10 min of rest. For all these parameters mean values were determined from two independent measurements and used for the study.

Laboratory data In the morning, after an overnight fast (8 h) and blood pressure measurement, a blood sample was withdrawn. Serum frozen samples (20  C) were shipped to the Laboratory of Molecular Medicine of the University of Rome ‘‘Tor Vergata’’ and the measurements of insulin, triglycerides (TG), total cholesterol (TC) and HDL-cholesterol (HDL-C) concentrations were assessed by standard immuno-enzymatic methods

‘‘The Linosa Study’’: MetS in a genetic isolate

457

(CV between 1.9 and 3.1%). Blood glucose was measured by glucoseeoxidase method on a Beckman glucose analyzer (Beckman Instruments, Fullerton, CA, USA). All these parameters were assessed twice and their means were used for the study. Finally, a total of 12 ml of blood was obtained from each participant, providing an average of 400 mg of DNA.

Diagnostic criteria The MetS was defined according to NCEP/ATP III diagnostic criteria [23] as the presence of three or more of the following features: hypertriglyceridaemia (HTR): >1.7 mmol/l; ‘‘low HDL’’: <1.04 mmol/l in men or <1.3 mmol/l in women; hypertension (HT): SBP > 130 mmHg or DBP > 85 mmHg; hyperglycaemia (HG): >6.1 mmol/l and abdominal obesity (cOB): WC > 88 cm in women and >102 cm in men. Participants who were taking antihypertensive (17.1%), antidiabetic (insulin or oral agents) (10.2%) or lipid lowering drugs (5.2%) were defined as hypertensive, hyperglycaemic or hyperlipidaemic, respectively. The presence of insulin resistance (IR) was assessed by homeostasis model assessment (HOMA-IR) [24]. The lower limit of the top quintile of HOMA-IR distribution (i.e., 2.35) in a sub-group of non-obese subjects from the Island with no metabolic disorders (n Z 76) was chosen as the threshold for IR in the LiS cohort, as suggested by the WHO [25]. HOMA-IR was calculated only among non-diabetic individuals (n Z 270), according to the WHO guidelines [25].

Statistical analysis Statistical analysis was performed with the SPSS 13.0 software (SPSS, Chicago, USA). Data from subjects on medications were excluded from the descriptive statistics reported in Table 1 as well as from the correlation analysis. Mean differences within generations between sexes were assessed for statistical significance using a ManneWhitney

Table 1

U-test (skewed variables). A generalized linear model was used to assess age-adjusted mean differences for metabolic risk factors between generations. Partial correlation coefficients (Pearson) adjusted for age and sex were used to compare the relationships between single MetS components. In this correlation analysis insulin, TG, HOMA-IR, SBP and DBP values were log-transformed. For all these analysis a p-value < 0.05 was considered statistically significant. Genetic analysis was carried out using SOLAR (Sequential Oligogenic Linkage Analysis Routines) software package to assess heritability estimates for the MetS and its dichotomized traits as defined in the ATP III. Furthermore, we estimated heritability for each continuous component of the MetS with adjustment for age, sex and concomitant medications to reduce the loss of information due to dichotomization of discrete traits. A total of 51 families (n Z 293), ranging in size from 3 to 10 individuals, were included in the estimation. Heritability (h2) of continuous variables was calculated using a standard quantitative genetic variance-components model implemented in SOLAR. With this approach, the maximum-likelihood estimation is applied to a model that incorporates fixed covariate effects, additive genetic effects and residual error, which are assumed to be normally distributed and mutually independent [26]; h2 is defined as the proportion of phenotypic variance that is attributable to additive genetic causes after accounting for covariates. The heritabilities of discrete traits were assessed using a threshold model in SOLAR. The method assumes that an individual belongs to a specific affection status if an underlying genetically determined risk (i.e., liability) exceeds a certain cut-off point [27]. The null hypothesis of no genetic effect (h2 Z 0) is tested by comparing the likelihood of a restricted model, where h2 is constrained to zero, with a general model in which the same parameter is estimated. Evidence of a non-zero estimate for a given parameter was considered statistically significant at p-value < 0.05.

Main clinical features of the LiS group. Parents (n Z 123)

Age (years) BMI (kg/m2) WC (cm) SBP (mmHg) DBP (mmHg) Glucose (mmol/l) Insulin (mU/ml) Homa-IR TC (mmol/l) HDL-C (mmol/l) TG (mmol/l)

Offsprings (n Z 170)

Females (n Z 65)

Males (n Z 58)

Females (n Z 83)

Males (n Z 87)

60.5  10.7 29.0  5.3 a,b 96.8  10.6a 142.8  18.5a 83.6  9.5a 5.94  1.83a 18.4  16.6a 5.08  8.59a,b 5.75  1.28a,b 1.01  0.21a 1.83  0.89a,b

65.8  11.2* 26.7  3.6* 97.3  10.5 140.8  17.7a,b 83.4  9.6a 5.67  1.55a 13.8  15.6*** 3.73  5.03a 5.49  0.86b 0.95  0.27b 1.72  0.92

31.7  9.6 25.2  4.5 84.8  12.4 118.8  17.1 73.6  10.9 4.78  0.67 13.1  10.0 2.73  2.51 5.13  0.99 1.11  0.26 1.10  0.61

34.4  9.9 26.3  3.8* 94.2  9.9*** 124.3  12.5* 76.9  9.3 4.82  0.78 10.5  8.8* 2.33  2.28 5.21  0.90 0.91  0.28*** 1.61  1.11**

Abbreviations are as described in the text. Data are presented as means  SD (n Z 293). *p < 0.05, **p < 0.01, ***p < 0.001 significant difference within generations between sexes (ManneWhitney U-Test). a p < 0.05 significant difference within sex between generations. b p < 0.05 after adjustment for age.

458

A. Bellia et al.

Results The major clinical and metabolic features of the 293 subjects from the LiS group, stratified by generation and sex, are given in Table 1. The mean  SD age for parents and offspring was 63.1  10.9 and 33.0  9.7 years, respectively, with a range of 18e92 years. In the parent group, females were younger than males and had a higher BMI, while no significant differences between genders were detected in WC and HDL-C mean values. Both in parents and in offsprings, females had higher insulin mean values, but no differences were observed in blood glucose concentrations and HOMA-IR. Conversely, in the offsprings group, BMI, WC, SBP and TG mean values were significantly lower in females, while HDL-C concentration was higher. Finally, when comparing different generations, females showed significant differences (p Z 0.05) in all metabolic variables with increasing age, while male offsprings did not differ from parents in BMI, WC, insulin levels and TG. According to NCEP/ATP III [23], 29.9% of individuals of the LiS group were classified as having the MetS (Fig. 1). These subjects were generally older and had higher mean BMI, WC and HOMA-IR values compared to those without the MetS (data not shown). Among MetS single components, prevalence rates were found to vary considerably, ranging from 20.3% for HG to 56.6% for ‘‘low HDL’’. cOB, HT and HTR accounted for 34.9%, 43.4% and 29.9%, respectively. According to HOMA-IR, nearly 40% of non-diabetic individuals were insulin-resistant while remarkably only 12.8% of the sample was found to be normal-weight and free of the above-mentioned metabolic disorders. Regarding the heterogeneity of the MetS phenotypes, the different combinations of NCEP/ATP III components showed a total of 16 phenotype clusters. Up to 54% of all MetS cases could be summarized into three clusters, each including cOB and ‘‘low HDL’’. The highest prevalence (23%) was observed for a phenotype combining elevated WC, TG at concomitantly low HDL-C. By contrast, the lowest prevalence of only 2% was found for a phenotype combining elevated blood pressure, glucose levels and low HDL-C with a normal WC. 56,6

43,4 34,9

%

29,9

37,6

29,9 20,3

MetS

HG

HTR

Low HDL

HT

cOB

IR*

Figure 1 Prevalence of the NCEP-ATP III MetS, related traits and insulin resistance. Abbreviations are as described in the text. Data are prevalence rates (n Z 293). *HOMA-IR > 2.35 (among non-diabetic subjects, n Z 270).

The prevalence for the remaining 12 MetS subtypes ranged between 3 and 10%. To determine the most prevalent contributions of combinations of the traits for MetS, we calculated Pearson partial correlations between single MetS components, adjusted for age and gender (Table 2). According to the above-reported observation about most prevalent MetS subtypes in the LiS group, WC showed a strong inverse correlation with HDL-C (0.30, p < 0.001) and a positive correlation with logTG (0.22, p < 0.001). Moreover, WC appeared significantly related with all NCEP/ATP III components and even with fasting insulin levels and HOMAIR. By contrast, glucose level and blood pressure components showed somewhat lower correlations with other MetS traits. As shown in Table 3, the MetS per se had a heritability (h2) of 27% (p Z 0.0012), while the variation explained by included covariates (i.e., age and sex) accounted for 12%. Among the 16 MetS subtypes, the most prevalent phenotype consisting of a combination of cOB, HTR and ‘‘low HDL’’, showed a remarkable h2 of 31% (p Z 0.002). When the single MetS components were treated as continuous traits, the estimates of heritability varied from 12% for DBP to 54% for HDL-C after accounting for relative covariates. Among NCEP/ATP III dichotomized traits, age- and sex-adjusted h2 values ranged from low (0.10 for HG, p Z NS) to a particularly large estimate of 0.49 (p < 0.001) for low HDL-C levels. IR showed appreciable heritability when entered as a discrete trait (h2 Z 0.32, p < 0.001) and even as HOMA-IR quantitative values (h2 Z 0.38, p < 0.001). We did not enter ongoing medications when analyzing discrete traits because patients on medications were reported as affected.

Discussion The ‘‘Linosa Study’’ was planned in order to verify the possibility of a further genetic analysis for complex diseases, such as the MetS, taking advantage of distinctive features of this unique islander and isolated population. Thus, we provide epidemiological and heritability data about the MetS and its related traits according to NCEP/ATP III criteria. In the LiS group, the overall prevalence of the MetS appeared fairly high (29.9%), especially when compared with data reported from NHANES III for a U.S. population [15] and from other Caucasian populations in Europe [17,19,29]. Furthermore, we found remarkable heritability estimates for the MetS per se (27%) and some of its individual components. The current definitions of the MetS are based upon different component traits; thus, patients with the MetS represent a heterogeneous group comprising multiple subtypes of the syndrome. It is unclear whether insulin resistance or visceral adiposity is the central feature linking together different components of the syndrome. Furthermore, it is unclear if the co-occurrence of the MetS traits may be explained by familial genetic factors (common genetic causes in close linkage or a shared major gene effect) or environmental effects such as lifestyle factors including nutrition and physical activity [28]. Considering this complexity, due to geneeenvironment interaction, the opportunity of studying an isolated population, with high

‘‘The Linosa Study’’: MetS in a genetic isolate Table 2

459

Partial correlation (Pearson), adjusted for age and sex, between NCEP-ATP III metabolic syndrome components. WC

WC

logSBP

0.19**

e

logSBP

0.19** 0.28*** 0.15* 0.37*** 0.32*** 0.30*** 0.22***

logDBP

Glucose loginsulin logHOMA-IR HDL-C logTG

logDBP

0.28*** 0.49***

e 0.49*** 0.19** 0.15* 0.17** 0.10 0.11

e 0.09 0.17** 0.15* 0.16* 0.17*

Glucose

logInsulin

logHOMA-IR

HDL-C

logTG

0.15* 0.19** 0.09 e 0.43*** NA 0.24*** 0.10

0.37*** 0.15* 0.17** 0.43*** e NA 0.22*** 0.37***

0.32*** 0.17** 0.15* NA NA e 0.21*** 0.25***

0.30*** 0.10 0.16* 0.24*** 0.22*** 0.21*** e 0.29***

0.22*** 0.11 0.17** 0.10 0.37*** 0.25*** 0.29*** e

Abbreviations are as described in the text. Skewed variables were log-transformed. *, **, ***significant correlation at *p < 0.05, **p < 0.01 and ***p < 0.001. NA Z not applicable.

consanguinity and strong founder effects, represented a potentially powerful strategy to identify new genetic variants. Thus, given these peculiar features, we decided to focus on the Linosa Island population. We restricted the analysis to the adult population (older than 18 years) given the lack of general consensus about MetS diagnostic criteria for children. Unexpectedly, given the mean age of the study group (48  18 years), the overall prevalence of the MetS in the LiS cohort was remarkably high. About one-third (29.9%) of the study group, who consisted of up to 70% of the Linosa population as a whole, met diagnostic criteria for the syndrome. This prevalence data is higher than those

reported in other industrialized Northern Italian regions [19,29] and in the NHANES III for a US Caucasian population [15]. The discrepancy, given the theoretically ‘‘low risk’’ environment in which the Linosa population lives, could not be attributed to differences in ethnicity or lifestyle and might warrant further genetic analysis. Moreover, despite remarkable prevalence rates for each MetS single trait (ranging from 20% for hyperglycaemia to 56% for ‘‘low HDL’’), the greatest contribution to overall MetS prevalence was driven by the co-occurrence of three major components: central obesity, low HDL and hypertriglyceridaemia. The observation of this clustering is in

Table 3 Heritability estimates (h2) for the NCEP-ATP III MetS, its traits and insulin resistance treated as continuous and discrete traits. h2 (SE)

p-value

Covariates

Variation explained by covariates

0.27 (0.11)

0.0012

Age, sex

0.12

0.27 (0.10) 0.38 (0.08)

0.002 <0.001

Age, sex Age, sex

0.11 0.09

Hyperglycemiaa Glucose (mmol/l)b

0.10 (0.09) 0.15 (0.04)

NS NS

Age, sex Age, sex, hypoglycemic treatment

0.36 0.31

Hypertriglyceridemiaa b logTG (mmol/l)

0.15 (0.10) 0.17 (0.07)

0.034 0.027

Age, sex Age, sex, lipid-lowering treatment

0.12 0.10

‘‘Low HDL’’a HDL-C (mmol/l)b

0.49 (0.17) 0.54 (0.12)

<0.001 <0.001

Age, sex Age, sex, lipid-lowering treatment

0.04 0.05

Insulin resistancea b logHOMA-IR

0.32 (0.12) 0.38 (0.09)

<0.001 <0.001

Age, sex Age, sex

0.27 0.20

Hypertensiona b logSBP (mmHg)

0.11 (0.04) 0.16 (0.07)

NS NS

0.42 0.40

(mmHg)b

0.12 (0.09)

NS

Age, sex Age, sex, antihypertensive treatment Age, sex, antihypertensive treatment

Metabolic syndromea Central obesity WC (cm)b

logDBP

a

Abbreviations are as described in the text. Skewed variables were log-transformed. a Treated as discrete trait. b Treated as continuous trait.

0.43

460 line with previous findings [8] and could support the observation that single MetS traits occur together more often than expected by chance [2], encouraging investigation into a common genetic basis linking together the single components of the syndrome. Furthermore, by heritability estimation, we evaluated the contributions of genetic factors to the phenotypic variability of the MetS and its traits. As reported above, in the LiS group the MetS per se showed a heritability of 27%. This takes into account the contributions of age and sex as covariates (12%, respectively), leading to the observation that, in the LiS group, nearly one-third of the familial aggregation of the MetS is attributable to genetic factors. In addition, among most prevalent MetS subtypes, the clustering of central obesity, low HDL and hypertriglyceridaemia showed an even higher heritability (31%) after accounting for the same covariates. This data might suggest that a significant portion of the observed variance in these metabolic abnormalities is due to genes that act through potentially related pathways. Among the other NCEP/ATP III MetS components, HDL-cholesterol showed the strongest heritability (nearly a half of the phenotypic variance is attributable to genetic factors); conversely, blood pressure and glucose had the lowest heritability estimation, probably due to the young mean age (33 years) of subjects in the offsprings group and consequently a low prevalence of these disorders. This hypothesis is further confirmed by the remarkable covariate effect observed for hyperglycaemia and hypertension (36% and 42%, respectively, when considered as discrete traits). A number of data have been published regarding MetS heritability among different ethnic groups [30,31]. Recently, the ERF study reported prevalence and heritability of the MetS in a Dutch isolated population [32]. According to NCEP/ATP III criteria, prevalence of the MetS in the ERF study was a little lower than reported in our study group (24.7% vs 29.9%, respectively), with a comparable sample mean age (48 years). Furthermore, heritability estimate for the MetS appears more relevant in the LiS group. Regardless of potential genetic effects that could be different in the two studied populations, the discrepancy might be attributable to a number of elements such as different sample sizes, different structure of pedigrees (nuclear families or more extended pedigrees) and covariates included in the analysis; for example, we decided to adjust the model for medication status when analyzing continuous variables, while we did not consider inbreeding a coefficient among covariates. As such, we did not include physical activity in the model given the incomplete data collection that could have confounded analysis results. We recognize that the challenge of studying this particular population has some limitations and strengths. First, the sample size could seem an issue for both heritability estimates and association studies; this observation is true for those traits e i.e., hyperglycaemia, and hypertension e unlikely to be detected in young subjects. Indeed, these traits showed a low prevalence when considering the offspring generation group (mean age of 33 years) and had no appreciable heritability when entered as either continuous or discrete traits. By contrast, analysis of other MetS components, such as waist circumference or HDL-C, which can occur even in a young age showed

A. Bellia et al. appreciable heritability; thus, traits chosen for the analysis, either continuous or discrete, represent a crucial point. Furthermore, genetically isolated groups, even if small, can be more useful than a large population in identifying rare variants with modest effects, given the increased chance of loss or fixation of some genetic variants, potentially leading to a lower/higher genetic diversity. In conclusion, these findings about prevalence and heritability of the MetS in the LiS group might be of interest in order to carry out a genome-wide linkage study, given the potential advantage of analyzing pedigrees in an isolated small population.

References [1] Girman CJ, Rhodes T, Mercuri M, Pyo ¨ra ¨la ¨ K, Kjekshus J, Pedersen TR, et al. The metabolic syndrome and risk of major coronary events in the Scandinavian Simvastatin Survival Study (4S) and the Air Force/Texas Coronary Atherosclerosis Prevention Study (AFCAPS/TexCAPS). Am J Cardiol 2004;93: 136e41. [2] Hanson RL, Imperatore G, Bennett PH, Knowler WC. Components of the ‘‘metabolic syndrome’’ and incidence of type 2 diabetes. Diabetes 2002;51:3120e7. [3] Kelishadi R, Gouya MM, Adeli K, Ardalan G, Gheiratmand R, Majdzadeh R, et al. Factors associated with the metabolic syndrome in a national sample of youths: CASPIAN Study. Nutr Metab Cardiovasc Dis 2008;18:461e70. [4] Grundy SM, Hansen B, Smith Jr SC, Cleeman JI, Kahn RA. Clinical management of metabolic syndrome: report of the American Heart Association/National Heart, Lung, and Blood Institute/American Diabetes Association conference on scientific issues related to management. Circulation 2004;109: 551e6. [5] De Pergola G, Pannacciulli N, Coviello N, Scarangella A, Di Roma P, Caringella M, et al. sP-selectin plasma levels in obesity: association with insulin resistance and related metabolic and prothrombotic factors. Nutr Metab Cardiovasc Dis 2008;18:227e32. [6] Reaven GM. Banting lecture 1988. Role of insulin resistance in human disease. Diabetes 1988;37:1595e607. [7] DeFronzo RA, Ferrannini E. Insulin resistance. A multifaceted syndrome responsible for NIDDM, obesity, hypertension, dyslipidemia, and atherosclerotic cardiovascular disease. Diabetes Care 1991;14:173e94. [8] Lemieux I, Pascot A, Couillard C, Lamarche B, Tchernof A, Alme ´ras N, et al. Hypertriglyceridemic waist: a marker of the atherogenic metabolic triad (hyperinsulinemia; hyperapolipoprotein B; small, dense LDL) in men? Circulation 2000; 102:179e84. [9] Maison P, Byrne CD, Hales CN, Day NE, Wareham NJ. Do different dimensions of the metabolic syndrome change together over time? Evidence supporting obesity as the central feature. Diabetes Care 2001;24:1758e63. [10] Branth S, Ronquist G, Stridsberg M, Hambraeus L, Kindgren E, Olsson R, et al. Development of abdominal fat and incipient metabolic syndrome in young healthy men exposed to long-term stress. Nutr Metab Cardiovasc Dis 2007; 17:427e35. [11] Hong Y, Pedersen NL, Brismar K, de Faire U. Genetic and environmental architecture of the features of the insulinresistance syndrome. Am J Hum Genet 1997;60:143e52. [12] Ruderman N, Chisholm D, Pi-Sunyer X, Schneider S. The metabolically obese, normal-weight individual revisited. Diabetes 1998;47:699e713.

‘‘The Linosa Study’’: MetS in a genetic isolate [13] Garg A, Misra A. Lipodystrophies: rare disorders causing metabolic syndrome. Endocrinol Metab Clin North Am 2004; 33:305e31. [14] Tassi V, Coco A, Padovano L, De Bonis C, De Cosmo S, Trischitta V. Akt2 gene common allelic variants in insulin resistance and the metabolic syndrome. Nutr Metab Cardiovasc Dis 2008;18:263e70. [15] 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: 356e9. [16] Meigs JB, Wilson PW, Nathan DM, D’Agostino Sr RB, Williams K, Haffner SM. Prevalence and characteristics of the metabolic syndrome in the San Antonio Heart and Framingham Offspring Studies. Diabetes 2003;52:2160e7. [17] Lakka HM, Laaksonen DE, Lakka TA, Niskanen LK, Kumpusalo E, Tuomilehto J, et al. The metabolic syndrome and total and cardiovascular disease mortality in middle-aged men. JAMA 2002;288:2709e16. [18] Bustos P, da Silva A, Amigo H, Bettiol H, Barbieri MA. Metabolic syndrome in young adults from two socioeconomic Latin American settings. Nutr Metab Cardiovasc Dis 2007;17:581e9. [19] Bonora E, Kiechl S, Willeit J, Oberhollenzer F, Egger G, Bonadonna RC, et al. Metabolic syndrome: epidemiology and more extensive phenotypic description. Cross-sectional data from the Bruneck Study. Int J Obes Relat Metab Disord 2003; 27:1283e9. [20] Loos RJ, Katzmarzyk PT, Rao DC, Rice T, Leon AS, HERITAGE Family Study, et al. Genome-wide linkage scan for the metabolic syndrome in the HERITAGE Family Study. J Clin Endocrinol Metab 2003;88:5935e43. [21] Langefeld CD, Wagenknecht LE, Rotter JI, Williams AH, Hokanson JE, Insulin Resistance Atherosclerosis Study Family Study, et al. Linkage of the metabolic syndrome to 1q23-q31 in Hispanic families: the Insulin Resistance Atherosclerosis Study Family Study. Diabetes 2004;53:1170e4. [22] Kissebah AH, Sonnenberg GE, Myklebust J, Goldstein M, Broman K, James RG, et al. Quantitative trait loci on chromosomes 3 and 17 influence phenotypes of the metabolic syndrome. PNAS 2000;97:14478e83.

461 [23] Expert Panel on the Detection Evaluation and Treatment of High Blood Cholesterol in Adults. Executive summary of the third report of the National Cholesterol Education Treatment Program (NCEP) Expert Panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). JAMA 2001;285:2486e97. [24] Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28:412e9. [25] Alberti KG, Zimmet PZ. 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;15:539e53. [26] Almasy L, Blangero J. Multipoint quantitative-trait linkage analysis in general pedigrees. Am J Hum Genet 1998;62: 1198e211. [27] Duggirala R, Williams JT, Williams-Blangero S, Blangero J. A variance component approach to dichotomous trait linkage analysis using a threshold model. Genet Epidemiol 1997;14: 987e92. [28] De Feo P, Stocchi V. Physical activity for the treatment and prevention of metabolic syndrome. Nutr Metab Cardiovasc Dis 2007;17:327e31. [29] Miccoli R, Bianchi C, Odoguardi L, Penno G, Caricato F, Giovannitti MG, et al. Prevalence of the metabolic syndrome among Italian adults according to ATP III definition. Nutr Metab Cardiovasc Dis 2005;15:250e4. [30] Austin MA, Edwards KL, McNeely MJ, Chandler WL, Leonetti DL, Talmud PJ, et al. Heritability of multivariate factors of the metabolic syndrome in nondiabetic Japanese Americans. Diabetes 2004;53:1166e9. [31] McQueen MB, Bertram L, Rimm EB, Blacker D, Santangelo SL. A QTL genome scan of the metabolic syndrome and its component traits. BMC Genet 2003;4(Suppl. 1):S96. [32] Henneman P, Aulchenko YS, Frants RR, van Dijk KW, Oostra BA, van Duijn CM. Prevalence and heritability of the metabolic syndrome and its individual components in a Dutch isolate: the Erasmus Rucphen Family study. J Med Genet 2008; 45:572e7.