Heritability of Determinants of the Metabolic Syndrome among Healthy Arabs of the Oman Family Study


Department of Biochemistry, College of Medicine and Health Sciences, Sultan Qaboos University, Al Khod, P.O. Box 35, Muscat 123, Sultanate of Oman. E-mail: bayoumi@squ.edu.om


The metabolic syndrome, as defined by the International Diabetes Federation, was investigated in five large, extended, highly consanguineous, healthy Omani Arab families of a total of 1277 individuals. Heritability (h2) of the phenotypic abnormalities that make up the syndrome and other related traits was estimated by variance decomposition method using SOLAR software. The overall prevalence of the syndrome was 23%. The prevalence of abnormalities making the syndrome in a descending order were: obligatory waist circumference, hypertension, raised fasting blood glucose, low serum high-density lipoprotein (HDL), and raised serum triglycerides (TGs). Highly significant, but widely spread, h2 values were obtained for: height (0.68), weight (0.68), BMI (0.68), serum HDL (0.63), serum leptin (0.55), percentage body fat (0.53), total serum cholesterol (0.53), fasting serum insulin (0.51), homeostasis model assessment-insulin resistance index (0.48), serum TG (0.43), waist circumference (0.40), diastolic blood pressure (0.38), and 2-hour glucose level (0.17), whereas for the metabolic syndrome itself, h2 was 0.38. The wide spread of h2 results (0.07 to 0.68) indicates that some determinants, such as weight, BMI, and HDL level, are under significant genetic influence among the Omani Arabs. Other determinants such as insulin resistance, abdominal obesity, diastolic blood pressure, and TG levels seem to be more environmentally driven.

The metabolic syndrome is an evolving entity of a cluster of conditions including abdominal obesity, insulin resistance (IR),1 glucose intolerance, type 2 diabetes, atherogenic dyslipidemia, endothelial dysfunction, prothrombotic and proinflammatory states, and hypertension and other cardiovascular disease (1, 2, 3). Overeating, a sedentary lifestyle, physical inactivity, and genetic susceptibility have been found to be the primary causes of the disease process. Individuals with this syndrome are more likely to die of coronary artery disease, heart failure, stroke, or renal failure. The prevalence of all-cause mortality due to the syndrome is, therefore, doubled. The syndrome has previously been clinically defined by the World Health Organization (4) and the Adult Treatment Protocol III of the U.S. National Cholesterol Education Program (5). Recently, the International Diabetes Federation (6) has used the term to define a patient with abdominal obesity and any two of the following four factors: elevated serum triglycerides (TGs), low-serum high-density lipoprotein (HDL) cholesterol, high blood pressure (BP), and high fasting plasma glucose (FPG). In Oman, the metabolic syndrome affects 21% of adults (7) tested in the same region as the present study. Our results are in agreement with those reported (7).

Clustering of the metabolic syndrome in families and the varying incidence of the syndrome among different ethnic groups suggest genetic susceptibility (8). Independently, the components of the syndrome also seem to have genetic determinants (9, 10). However, the study of the genetics of the metabolic syndrome is fraught with difficulties. The complexity of the interaction of genetic susceptibility and environmental determinants that shape the phenotype require different and novel approaches to the problem. We have, therefore, chosen a unique family model in an attempt to disentangle the genetic and environmental factors that make up this syndrome (11). Our Arab study population is a relatively homogeneous traditional group with a very large pedigree size (140 to 300) and very high degree of inbreeding and polygamy that has been subjected to an accelerated change in lifestyle over the past 30 years, resulting in a high prevalence of the metabolic syndrome (7). On testing this study population, very high statistical power was obtained for h2 of the determinants of the metabolic syndrome (11).


Apparently healthy adults, 16 to 80 years old, were sampled from five Omani Arab pedigrees (Table 1). The pedigrees consisted of 327, 160, 230, 279, and 281 individuals (a total of 1277). Although the total number of founders ranged between 70 and 100 in each pedigree, most of these were due to marriages outside the pedigrees. The 5th to 10th generation founders were usually very few and ranged between 3 and 17 individuals (Table 1). The rapid population growth made these pedigrees fairly young isolates of 7 to 12 generations each.

Table 1.  Pedigree data
Total closely related individuals508309511649464
Individuals tested327160230279281
Nuclear families108609810187
Mean inbreeding coefficient0.01060.00920.02060.01800.0183

Of the total examined, only 1198 individuals (536 men and 662 women) had complete sets of data. Anthropometric, biochemical, and BP characteristics of the whole population are listed in Table 2. Known gender differences in height, weight, HDL, TG, and BP were confirmed. Although men and women had the same BMI, women seemed to have a much higher percentage body fat, larger waist circumference (in relation to height and weight), and higher glucose intolerance.

Table 2.  Characteristics of the whole Omani Arab study group (n = 1198)
VariablesTotal (n = 1198) [mean (±SD)]Men (n = 536) [mean (±SD)]Women (n = 662) [mean (±SD)]p
  1. NS, not significant.

Age (years)33.8 (16.2)33.0 (17.0)34.5 (15.5)0.005
Height (cm)158.3 (9.3)165.8 (7.3)152.2 (5.5)<0.001
Weight (kg)62.9 (14.6)68.1 (14.4)58.7 (13.4)<0.001
BMI (kg/m2)25.1 (5.4)24.8 (5.0)25.7 (5.7)0.1
Body fat (%)23.5 (10.5)17.9 (8.4)28.1 (9.8)<0.001
Waist circumference (cm)81.0 (14.5)81.1 (14.3)80.0 (14.6)0.9
LDL (mM)3.2 (1.0)3.3 (1.1)3.2 (1.0)0.1
HDL (mM)1.1 (0.3)0.9 (0.2)1.2 (0.3)<0.001
TG (mM)1.1 (0.8)1.2 (0.8)1.0 (0.8)<0.001
FPG (mM)5.6 (1.5)5.6 (1.3)5.7 (1.6)0.3
PG2 (mM)6.8 (3.3)6.2 (2.7)7.3 (3.6)<0.001
Insulin 0 (mIU/L)5.6 (4.3)5.6 (4.1)5.7 (4.4)0.9
HOMA-IR1.39 (1.1)1.37 (1.01)1.41 (1.17)0.6
Hemoglobin A1c (%)5.5 (1.8)5.4 (1.0)5.6 (2.2)0.3
Leptin (ng/mL)26.9 (23.8)13.1 (13.6)38.1 (24.4)<0.001
Growth hormone (mIU/L)4.6 (9.8)2.0 (5.8)6.7 (11.7)<0.001
Daytime SBP (mm Hg)124.5 (14.2)129.4 (14.1)120.6 (12.9)<0.001
Daytime DBP (mm Hg)82.1 (10.3)83.4 (10.4)81.1 (10.2)<0.001

The prevalence of the syndrome was 23% according to International Diabetes Federation (IDF) criteria. This is similar to findings in the same population by Al Lawati et al. (7). Results are also not different from those in other ethnic groups (12, 13, 14, 15, 16, 17). However, women in this population seem to be more affected (n = 188) than men (n = 88): a ratio of 2.14:1.00. The waist circumference in these women appears to correlate directly with parity: the higher the parity, the larger the waist. The average parity in this population is 6 to 7. Thus, the greater proportion of women with the syndrome appears to be directly related to the much higher parity among Omani Arab women.

Prevalence of Morbidities

The prevalences of morbidities defining the syndrome, in descending order, in the whole study population (n = 1198) were: low serum HDL, raised BP, large waist circumference, raised FPG, and raised serum TG. Of the study population, 85% had one or more of the syndrome morbidities. Men were characterized by higher BP and low HDL, whereas women had large waist circumferences.

h2 of the Phenotypic Abnormalities

The relative pairs of individuals used in the quantitative genetic analysis in the five pedigrees were 23,253 pairs. Of those pairs, there were 2482 parent-offspring, 1278 siblings, 3774 grandparent-grandchild, 2815 avuncular, 2610 first cousins, 4363 second cousins, and the remaining with more complex relationships. This large number of pairs is due to the very high degree of inbreeding and complexity of pedigrees. The h2 estimate of the IDF-defined metabolic syndrome itself in this Arab population was 0.378 ± 0.12 (p < 1.2 × 10−5). The h2 estimates of the IDF-defined determinants of the syndrome and other related traits (Table 3) were: height (0.68), weight (0.68), BMI (0.68), HDL level (0.63), leptin (0.55), percentage body fat (0.53), total cholesterol level (0.53), fasting insulin level (insulin 0; 0.51), homeostasis model assessment (HOMA)-IR index (0.48), TG level (0.43), diastolic BP (DBP; 0.38), waist circumference (0.40), 2-hour glucose level (0.17), hemoglobin A1c (0.10), and FPG (0.07). Very high levels of statistical significance were obtained with even low heritabilities due to the unique structure of these Arab pedigrees.

Table 3.  h2 Estimates of anthropometric, biochemical, and BP parameters among 1198 Omani Arabs
Traith2SEpProportion of variance attributed to covariate effects
  • *

    h2 Estimates were adjusted to covariate effects of age, age2, sex, age × sex, and age2 × sex.

Height (cm)0.680.061.5 × 10−370.61
Weight (kg)0.680.052.1 × 10−480.23
BMI (kg/m2)0.680.051.03 × 10−510.16
Body fat (%)0.530.061.0 × 10−360.43
Waist circumference (cm)0.400.063.2 × 10−210.31
Cholesterol (mM)0.530.065.8 × 10−340.23
LDL (mM)0.480.061.5 × 10−320.17
HDL (mM)0.630.066.9 × 10−440.20
TG (mM)0.430.065.3 × 10−250.22
FPG (mM)
PG2 (mM)0.170.055 × 10−70.18
Insulin 0 (mIU/L)0.510.051.3 × 10−430.02
HOMA-IR0.480.065.8 × 10−370.03
HbA1C (%) × 10−50.03
Leptin (ng/mL)0.550.064.1 × 10−400.35
Growth hormone (mIU/L) × 10−110.37
Daytime SBP (mm Hg) × 10−180.16
Daytime DBP (mm Hg)0.380.054.4 × 10−290.06

The h2 of the metabolic syndrome itself, as a threshold discrete trait, among these five Omani pedigrees (0.38) is similar to that obtained in other ethnic groups (12, 17). The h2 of the IDF determinants and other related phenotypes showed a wide spread (0.07 to 0.68). Some phenotypes, such as weight, BMI, HDL level, and IR, seem to be under considerable genetic influence, whereas other phenotypes such as TG, abdominal obesity, systolic BP (SBP), DBP, and FPG levels seem to be affected largely by unmeasured non-genetic factors, probably environmental. Among the definitive IDF metabolic syndrome determinants, heritabilities range between 0.63 and 0.07. HDL-cholesterol had the highest estimate (0.63), followed by TG (0.43), waist circumference (0.40), DBP (0.38), SBP (0.28), and least for FPG (0.07).

In general, our data are in agreement with previous studies, where h2 of determinants of the metabolic syndrome were estimated in whites (13, 16), Hispanics (17), Chinese, and Japanese (14, 15). h2 estimates in Omani Arabs seem closer to Hispanics than whites. No significant differences in h2 estimates were observed between Omani Arabs and other ethnic groups in all determinants and related phenotypes, except in FPG and plasma glucose 2 hours after 75-g oral glucose load (PG2), where lower estimates were obtained for the Omani Arabs (12, 13, 14, 15, 16, 17).

The Omani Arabs subjected to a rapid change in lifestyle seem to suffer from metabolic syndrome in a similar manner to other ethnic groups. However, the higher proportion of women affected seems to be due to the greater parity-related increase in waist circumference. The h2 estimates indicate that some determinants of the syndrome are under significant genetic influence among the Omani Arabs, which lends more credibility and support for future projects on association and gene mapping of the determinants of the metabolic syndrome among Arabs.

Research Methods and Procedures

Study Area and Population

The Interior Province of Oman, where the study was conducted, is 80 to 150 km south of the capital, Muscat. It is a mountainous region dotted by several oases in river beds where traditional agriculture of mainly date palm and subsistence farming and animal breeding have been practiced by successive Arab generations. Although the older generations are still working in traditional agriculture and animal breeding, the younger and educated men and women have taken light manual jobs such as light vehicle drivers and labor supervisors. Due to the 30-year oil boom, a dramatic change in the lifestyle of the population has been witnessed.


Five large, extended, and highly consanguineous families, each living in a separate village, were selected within a perimeter of 20 km around Nizwa. The number of adult subjects interviewed and tested in these five pedigrees was 1277, which represented roughly 10% to 15% of the total people in the five villages. Their ages ranged between 16 and 80 years. They all voluntarily took part in the study. An explained written consent form was obtained from each individual. The study has been approved by the Ethics and Research Committee of the College of Medicine, Sultan Qaboos University. All individuals appeared healthy and had no clinical complaints. Detailed inquiry revealed that they all seemed to be related to common ancestry, and intermarriage is common among them. Cousin marriages represent >35% of all marriages and a further 20% between tribal groupings (11). Polygamy is widely practiced, with some men marrying up to four wives. Family relationships were ascertained initially by local staff, volunteers, and elders from each village. A questionnaire was filled in for each subject, each of whom was given a unique identification number that was used for creating a master file for all tests and analyses.

Demographic Data Collection

A satellite Research Centre was established in Nizwa Polyclinic. This center was used for anthropometric measurements, sample collection, and BP studies. Height and weight were measured using standard methods. Waist circumference was measured by a soft tape at the largest circumference between the lowest rib and iliac crest. Body fat percentage was assessed using electrical impedance (Tanita, Tokyo, Japan). Biochemical tests included FBG and PG2, fasting and 2-hour serum insulin levels, fasting total serum cholesterol, low-density lipoprotein (LDL)- and HDL-cholesterol, and TGs. The HOMA was used to evaluate IR. HOMA is calculated as [FPG (millimolar) × fasting insulin (microunits per milliliter)/22.5]. All biochemical tests were done the same day of sample collection or frozen at −80 °C until done at the University Hospital laboratories using modern automated equipment (Synchron 7, Access II and Image; Beckman Coulter, Fullerton, CA). Quality is assured by participation in international and local quality control programs.

BP was measured by standard mercury sphygmomanometer. In addition, the means of 24-hour ambulatory BP were recorded (Schiller AG, Baar, Switzerland). Quality assurance of all work was ascertained by duplicate measurements and restricting parameters to maximally two matched observers for anthropometric, BP, and other measurements (11).


IDF criteria (6) were used for defining the metabolic syndrome. A person with the syndrome must have central obesity (defined as waist circumference ≥ 94 cm for men and ≥80 cm for women), plus any two of the following four factors: raised TG level > 1.7 mM or specific treatment for it; reduced HDL-cholesterol < 0.9 mM in men and <1.1 mM in women or specific treatment for it; raised SBP of ≥130 or DBP of ≥85 mm Hg or treatment of previously diagnosed hypertension; or raised FPG ≥ 5.6 mM or previously diagnosed type 2 diabetes.

The SPSS statistical package version 10.0 for personal computers (SPSS, Inc., Chicago, IL) was used for initial statistical analysis. Data are presented as average (± standard deviation) or standard error. The parameter distributions were tested by Kolmogorov-Smirnov criteria and were log-normalized if required.

Inbreeding level was calculated using the Quaas-Henderson algorithm (18). The inbreeding coefficient for each subject was calculated on the probability that two genes at the same locus drawn at random, one from each parent, will be identical by descent. The mean inbreeding coefficient was the average for each pedigree (19). To compute the inbreeding coefficients, PEDSYS (Pedigree Data Management System, version 2.0; Southwest Foundation for Biomedical Research, San Antonio, TX) was used.

h2 has two definitions. The first is a statistical definition, which defines h2 as the proportion of phenotypic variance attributable to genetic variance. The second definition is more “common sensical.” It defines h2 as the extent to which genetic individual differences contribute to phenotypic individual differences. h2 and Kullback-Leibler pseudo-R (2) for categorical variables were calculated for variability of the trait as explained by the independent variables. For calculating heritabilities, the metabolic syndrome determinants and related phenotypes were considered as quantitative discrete traits, whereas for the syndrome itself, the phenotype was considered as a threshold discrete trait. Computation was carried out using the SOLAR software package (Southwest Foundation for Biomedical Research) (20).


This work was supported by His Majesty Sultan Qaboos Strategic Research Trust Fund (Grant SR/MED/PHYS/04/01). We thank the Ministry of Health and the Wali (Governor) of Nizwa District for supporting the work. We also thank the nursing staff of the Nizwa Hypertension Project and the people of the Taimsa, Birkat Al-Moz, Radat Al-Busaidi, Karsha, and Farq villages for their continuous encouragement and participation. We also thank Jessy George and Taruna Dutt for manuscript preparation.


  • 1

    Nonstandard abbreviations: IR, insulin resistance; TG, triglyceride; HDL, high-density lipoprotein; BP, blood pressure; FPG, fasting plasma glucose; IDF, International Diabetes Federation; HOMA, homeostasis model assessment; DBP, diastolic BP; SBP, systolic BP; PG2, plasma glucose 2 hours after 75-g oral glucose load; LDL, low-density lipoprotein.

  • The costs of publication of this article were defrayed, in part, by the payment of page charges. This article must, therefore, be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.