Atherosclerosis 207 (2009) Contents lists available at ScienceDirect. Atherosclerosis

Transcription

1 Atherosclerosis 207 (2009) Contents lists available at ScienceDirect Atherosclerosis journal homepage: Levels of lipids and apolipoproteins in three cultures Amirreza Solhpour a, Sahar Parkhideh a, Nizal Sarrafzadegan b, Sedigheh Asgary b, Ken Williams c, Ingmar Jungner d, Are Aastveit e, Göran Walldius f, Allan Sniderman a, a Mike Rosenbloom Laboratory for Cardiovascular Research, McGill University Health Centre, Room H7.22, Royal Victoria Hospital, 687 Pine Avenue West, Montreal, Quebec H3A 1A1, Canada b Isfahan Cardiovascular Research Center and Isfahan University of Medical Sciences, Isfahan, Iran c KenAnCo Biostatistics, San Antonio, TX, USA d Department of Medicine, Clinical Epidemiology Unit, Karolinska Institutet, and CALAB Research, Stockholm, Sweden e Department of Chemistry, Biotechnology and Food Science, Norwegian University of Life Sciences, Aas, Norway f King Gustaf V Research Institute and Karolinska Institutet, Stockholm, Sweden article info abstract Article history: Received 26 May 2009 Received in revised form 31 August 2009 Accepted 2 September 2009 Available online 6 September 2009 Keywords: Lipids Apolipoproteins Vascular Disease Risk Cultures Objective: A complete lipoprotein profile requires measurement of the major plasma apolipoproteins apob and apoa-i in addition to measurement of the major plasma and lipoprotein lipids. The objective of this study was to demonstrate the additional information that can be acquired by comparing the major plasma lipids and apolipoproteins amongst adult male and female Swedes, Iranians and Americans. Methods: Data on Iranians were derived from the Isfahan Healthy Heart Program, a prospective community-based sample of 12,103 individuals >20 years of age. Data on Swedes were derived from the AMORIS study, a prospective epidemiological study of 173,629 subjects. Data on Americans were derived from the NHANES III, data bank, which is designed to be representative of the adult American population. Lipids were measured by conventional methods. ApoB and apoa-i were measured by IFCC/WHO standardized methods. Results and Conclusions: There is a complex pattern of differences amongst the cultures. There are also important similarities in the differences between the genders. Swedes have the highest levels of LDL-C (3.77 mmol/l), apob (1.27 g/l) and HDL-C (1.53 mmol/l) but the lowest levels of triglyceride (1.47 mmol/l for the Swedes vs mmol/l and 1.93 mmol/l for the Americans and Iranians, respectively, all p < 0.001). Americans have higher levels of LDL-C than Iranians (3.26 mmol/l vs mmol/l, p < 0.001) but lower levels of apob (1.04 g/l vs g/l, p < 0.001). The absolute values for both HDL-C and apoa-i are higher in females than males of all three cultures throughout the population distribution. The levels of the atherogenic lipoproteins peak between 30 and 40 in males but continue to rise in females. The apob/apoa-i ratio is highest in the Swedes (0.92, p < 0.001) but similar in the Americans and Iranians (0.81 and 0.80, respectively, p NS). By contrast, the TC/HDL-C ratio is highest in the Iranians, intermediate in the Americans and lowest in the Swedes (4.41 vs vs. 4.22, all p < 0.001). These data provide further evidence that complete characterization of lipoproteins requires measurement of apob and apoa-i as well as lipoprotein lipids and that the changes in plasma lipoproteins over time differ between the genders Elsevier Ireland Ltd. All rights reserved. 1. Introduction INTERHEART, the first study to systematically examine the determinants of vascular disease in all the major regions of the world, demonstrated that the same nine modifiable factors accounted for virtually all the population-attributable risk of acute myocardial infarction (AMI) [1]. Of these, the balance of the Corresponding author. Tel.: x34637; fax: address: allansniderman@hotmail.com (A. Sniderman). pro-atherogenic and anti-atherogenic lipoproteins in plasma, as estimated by the apolipoprotein B/apolipoprotein A I (apob/apoa- I) ratio, was the most important. Nevertheless, this ratio accounted for just half the risk, further evidence that it is the overall risk profile that determines the overall risk [1]. Cholesterol has been the conventional index to measure the concentrations of low density lipoprotein (LDL) and high density lipoprotein (HDL) particles in plasma, whereas triglycerides have been used to estimate the concentration of very low density lipoprotein (VLDL). But there are many disadvantages to this approach. In particular, when small, dense, cholesterol-depleted LDL particles predominate, as they /$ see front matter 2009 Elsevier Ireland Ltd. All rights reserved. doi: /j.atherosclerosis

2 A. Solhpour et al. / Atherosclerosis 207 (2009) often do in patients at high risk of vascular disease, LDL-C necessarily underestimates LDL particle number [2]. Alternatively, since each VLDL and LDL particle contains one molecule of apob, total atherogenic particle number can be accurately estimated using standardized methods by measuring plasma apob [3]. The INTERHEART study demonstrated that apob and apoa-i, either individually, or taken together as the apob/apoa- I ratio were substantially superior as risk markers of AMI than the conventional cholesterol indices [4]. The strength of INTER- HEART is the breadth of subjects that were studied and the fact that even though INTERHEART was case-control in design, its results are virtually identical to prospective longitudinal studies such as apolipoprotein-related mortality risk (AMORIS) [5]. While INTER- HEART demonstrated a common pool of modifiable risk factors accounted for virtually all the risk, this does not necessarily mean these factors are present to the same degree amongst all the peoples of the world. For example, vascular disease occurred earlier in South Asians than in the peoples from other countries, was more common in South Asian females than females elsewhere, and was more commonly associated with an increased apob/apoa-i ratio, smoking, and an increased waist to hip ratio (WHR) [6]. Thus, while the same risk factors operate everywhere, their prevalence and the age at which they appear, and therefore their individual prominence as drivers of disease, can vary substantially. If we are to understand the differences in the incidence and prevalence of vascular disease amongst the peoples of the world, we must understand the ways in which these risk factors differ as well as the ways in which they are similar. Accordingly, the present study examines plasma lipids as well as apob and apoa-i in three cultures: Sweden, Iran, and the United States. Our objective was to see if differences existed amongst these three cultures, if they were present in both genders and had similar patterns of change across our lifetimes, and whether apolipoproteins and lipids mirrored these similarities or differences equally. 2. Methods Data were obtained from each of three groups as follows: 2.1. Iranians Subjects were identified within the framework of the Isfahan Healthy Heart Program (IHHP), a comprehensive integrated community-based program to control and prevent cardiovascular diseases, started in The IHHP population sampling procedures have been described in detail previously [7]. Individuals were sampled with the use of a highly stratified, multistage probability design. The study participants comprised 5895 men (mean age 39.8 years) and 6208 women (mean age 39.4 years). We limited the present analysis to individuals aged 20 (This excluded 411 participants who were under 20 years old). Apolipoproteins were measured on 1690 subjects (761 males and 929 females). The study was approved by the Ethics Review Board of Isfahan Cardiovascular Research Center. Serum samples were obtained by venipuncture after a h fast. Fresh blood samples were used to measure Total Cholesterol (TC), Triglyceride (TG) and HDL-C and measurements of apob and apoa-i were carried out on plasma that had been frozen ( 78 C). Total serum cholesterol and triglyceride were determined by standard enzymatic method using special kits (Immunodiagnostic, Germany) in an Elan 2000 auto-analyzer. (Eppendorf, Germany). HDL-C was measured enzymatically after precipitating the other lipoproteins with dextran sulphate magnesium chloride. LDL-C was calculated from the three primary measurements with the use of the Friedewald equation [8]: Direct measurement of LDL-C was performed with a turbidimetric method for those with TG levels higher than 4.5 mmol/dl. Concentrations of apoa-i and apob were determined by kinetic immunoturbidimetric methods on a Hitachi auto-analyzer, model 705 as described in detail elsewhere [9]. All the analyses were performed at the laboratory of Isfahan Cardiovascular Research Center. For quality control, the central laboratory meets the criteria of the Ministry of Health in Iran. An external standardization for lipids was performed with the central laboratory of the St. Rafael University Hospital of Leuven in Belgium. The measurements of apob and apoa-i were standardized by the North- West Lipid Research Laboratories, Seattle, Washington, based on the procedures of the World Health Organization-International Federation of Clinical Chemistry (WHO-IFCC) [3]. The coefficients of variation for TG and TC were <8% and HDL-C <6%. The coefficients of variation were 1.5% and 1.4% for apob and apoa-i, respectively Americans Data describing the American population was derived from the National Health and Nutrition Examination Survey (NHANES) III data bank. NHANES has a complex sample design which was designed to develop statistics representative of the US noninstitutionalized civilian population from 1988 to Detailed methods used in NHANES III are published elsewhere and are available for public access on the Internet [10]. Population standard deviations were estimated by weighting each individual by the population weight multiplied by the sample size/the sum of population weights as suggested by Harrell [11]. The present analysis was restricted to persons aged 20 years. LDL-C was calculated only on subjects who were randomly selected to come to morning sessions of the clinical examination after fasting for at least 8 h. For consistency all cholesterol and triglyceride analyses were restricted to this subgroup (n = 8515; 3991 men, 4524 women). Apolipoprotein measures were only conducted on subjects who were examined during phase 1 ( , n = 8577; 4286 men, 4291 women). The analytical methods are described in detail elsewhere. Specifically, the measurements of apoa-i and apob were compliant with the WHO-IFCC apolipoprotein program as previously documented [3]. The coefficient of variation for TC was <3%, TG < 4%, HDL-C < 4%, apoa-i and apob < 7% Swedes The AMORIS study was the source of data for the Swedish population [5]. The AMORIS population and analytical procedures have been described in detail previously [5,12]. Subjects were recruited consecutively at the time of a health check up. The present analysis was limited to individual s aged 20. This resulted in an analytic sample of 173,629 subjects; 97,796 men and 75,833 women. Measurement of apoa-i and apob values also conformed to the WHO-IFCC international criteria [3,12]. The coefficient of variation of measurement of TG, apob, apoa-i were all <7% and TC < 3% Statistical analysis Descriptive analysis included the estimation of mean values and standard deviation (SD) for continuous variables. These values were rounded at nearest integer or first decimal. All statistical analysis for the Iranian database was conducted using SPSS 15.0 for windows software (SPSS Inc., Chicago, IL, USA). For American data in NHANES III, SAS version 9.1 (Cary, NC, USA) SURVEY procedures were used. Significance of differences between three populations (Iran, Sweden and USA) was tested by one-way analysis of variance (ANOVA). We conducted pair-wise comparisons using the least significant difference (LSD) method with the help of Microsoft Office Excel We compared each parameter in males, females and

3 202 A. Solhpour et al. / Atherosclerosis 207 (2009) the total population separately. We considered a p value 0.05 statistically significant. 3. Results 3.1. TC, LDL-C, apob The number of subjects in each group along with their gender and age are listed in Table 1. There were significant differences in age amongst the three cultures with the Swedes being approximately 4 years older than the Americans who were, in turn, 4 years older than the Iranians. In both genders, TC was significantly higher in Swedes than in Americans or Iranians. The same pattern held for LDL-C in females, whereas in males there was a progressive gradient in values of LDL-C, the highest in Swedes, intermediate for Americans, and lowest in Iranians. On the other hand, while apob was highest in Swedes, in both males and females, the average values of apob were significantly higher in Iranians compared to Americans. Because of the difference in average age amongst the subjects in the three cultures, more detailed analyses of the distribution of values over the age distribution of the three populations were conducted. The range of values the 25th, 50th and 75th percentiles were determined and compared. These results confirm the overall relationships from the averages in that the rank ordering for all parameters were the same throughout the range of values. (On-line Figs. 1 3). The age analyses (see Figs. 1 3) also confirm that the differences in lipid and apolipoprotein levels amongst the three cultures are not due to the average difference in ages amongst them. ApoB and LDL-C are concordant in Swedes in that both are the highest in this culture but are discordant in Iranians and Americans. If LDL-C is the standard, Americans have a more pro-atherogenic profile than Iranians, whereas if apob is the standard, the reverse is the case. The pattern for triglycerides is very different: Values are lowest in the Swedes, intermediate in the Americans, and highest in Iranians. Iranian men and women, on average, had triglyceride values more than 10% higher than Swedes and the Americans. Swedish and American men had very similar range of triglyceride values, whereas American women had higher levels than Swedish women HDL-C and apoa-i The overall differences amongst HDL-C and apoa-i in the three cultures are listed in Table 1. In males, HDL-C is highest in Swedes whereas the values for HDL-C are very similar for Iranians and Americans. By contrast, the values for apoa-i differ little. The absolute values for both HDL-C and apoa-i are higher in females than males of all three cultures throughout the population distribution. Amongst females, HDL-C levels differ with levels being highest in Swedes and lowest in Iranians (Table 1). It is noteworthy that there is much less difference in the apoa-i results. The contrast between apoa-i and HDL-C is most prominent in the Iranian females with Fig. 1. Mean values of TC and TG in different age groups in three cultures. Panel A. TC male; Panel B. TG male; Panel C. TC female; Panel D. TG female

4 Table 1 Summary of the average results for males and females from each of the three countries. Measure Group A, Sweden mean (SD) Age mean (SD) n Group B, USA mean (SD) Age mean (SD) n Group C, Iran mean (SD) Age mean (SD) n p Value Group A vs. B p Value Group A vs. C p Value Group B vs. C TC (mmol/l) All 5.92(1.22) 48.5(13.6) (1.58) 44.6 (24.4) (1.3) 39.5 (14.7) <0.001 <0.001 <0.001 M 5.91(1.18) 47.3(12.7) (1.51) 41.2 (23.7) (1.28) 39.7 (15.0) 5775 <0.001 <0.001 <0.001 F 5.94(1.27) 50.1(14.6) (1.64) 47.3 (24.9) (1.30) 39.3 (14.3) 6110 <0.001 < TG (mmol/l) All 1.47(1.10) 48.5(13.6) (1.94) 44.6 (24.4) (1.25) 39.5 (14.7) <0.001 <0.001 <0.001 M 1.65(1.24) 47.3(12.7) (2.18) 41.2 (23.7) (1.36) 39.7 (15.0) <0.001 <0.001 F 1.23(0.84) 50.1(14.6) (1.69) 47.3 (24.9) (1.13) 39.3 (14.3) 6113 <0.001 <0.001 <0.001 LDL-C (mmol/l) All 3.77(1.08) 48.5(13.6) (0.96) 44.6 (24.4) (1.06) 39.2 (14.7) <0.001 <0.001 <0.001 M 3.81(1.05) 47.3(12.7) (0.94) 41.2 (23.7) (1.05) 39.5 (15.1) 5334 <0.001 <0.001 <0.001 F 3.72(1.12) 50.1(14.6.1) (0.98) 47.3 (24.9) (1.07) 39.0 (14.3) 5705 <0.001 < HDL-C (mmol/l) All 1.53(0.42) 48.5(13.6) (0.52) 44.6 (24.4) (0.28) 39.5 (14.7) <0.001 < M 1.41(0.38) 47.3(12.7) (0.49) 41.2 (23.7) (0.27) 39.7 (15.0) 5732 <0.001 < F 1.69(0.40) 50.1(14.6) (0.56) 47.3 (24.9) (0.28) 39.3 (14.3) 6039 <0.001 <0.001 <0.001 ApoB (g/l) All 1.27(0.36) 48.5(13.6) (0.27) 44.6 (24.4) (0.32) 38.4 (14.4) 1690 <0.001 <0.001 <0.001 M 1.31(0.36) 47.3(12.7) (0.25) 41.2 (23.7) (0.31) 39.2 (15.1) 761 <0.001 <0.001 <0.001 F 1.21(0.36) 50.1(14.6.1) (0.29) 47.3 (24.9) (0.32) 37.9 (13.8) 929 <0.001 <0.001 <0.001 ApoA-I (g/l) All 1.43(0.22) 48.5(13.6) (0.25) 44.6 (24.4) (0.43) 38.5 (14.4) <0.001 <0.001 M 1.36(0.21) 47.3(12.7) (0.22) 41.2 (23.7) (0.47) 39.2 (16.1) <0.001 <0.001 F 1.51(0.24) 50.1(14.6) (0.28) 47.3 (24.9) (0.40) 37.9 (13.8) <0.001 <0.001 ApoB/apoA-1 All 0.92(0.30) 48.5(13.6) (0.34) 44.6 (24.4) (0.21) 38.5 (14.4) 1690 <0.001 < M 0.99(0.31) 47.3(12.7) (0.35) 41.2 (23.7) (0.21) 37.9 (13.8) 761 <0.001 < F 0.82(0.29) 50.1(14.6) (0.32) 47.3 (24.9) (0.21) 39.2 (15.1) 929 <0.001 <0.001 <0.001 TC/HDL-C All 4.22(1.74) 48.5(13.6) (2.34) 44.6 (24.4) (1.33) 39.5 (14.7) <0.001 < M 4.59(1.96) 47.3(12.7) (2.49) 41.2 (23.7) (1.36) (15.0) 5681 <0.001 <0.001 <0.001 F 3.74(1.41) 50.1(14.6) (2.19) 47.3 (24.9) (1.30) 39.3 (14.4) 5995 <0.001 <0.001 <0.001 non-hdl-c/hdl-c All 3.22(1.74) 48.5(13.6) (2.34) 44.6 (24.4) (1.33) 39.5 (14.7) <0.001 < M 3.59(1.96) 47.3(12.7) (2.49) 41.2 (23.7) (1.36) 39.7 (15.0) 5681 <0.001 <0.001 <0.001 F 2.74(1.41) 50.1(14.6) (2.19) 47.3 (24.9) (1.30) 39.3 (14.4) 5995 <0.001 <0.001 <0.001 FBS (mg/dl) All 89.11(20.98) 48.5(13.6) (38.23) 44.6 (24.4) (32.13) 39.5 (14.7) <0.001 <0.001 <0.001 M 91.44(22.86) 47.3(12.7) (39.94) 41.2 (23.7) (30.12) 39.7 (15.0) 5769 <0.001 <0.001 <0.001 F 86.40(18.54) 50.1(14.6) (36.63) 47.3 (24.9) (33.93) 39.3 (14.3) 6130 <0.001 <0.001 <0.001 BMI (Kg/m 2 ) All 24.52(3.66) 48.5(13.6) (7.84) 44.6 (24.4) (4.65) 39.5 (14.6) <0.001 <0.001 <0.001 M 25.10(3.41) 47.3(12.7) (7.21) 41.2 (23.7) (4.09) 39.7 (15.0) 5851 <0.001 <0.001 <0.001 F 23.76(3.98) 50.1(14.6) (8.37) 47.3 (24.9) (5.13) 39.3 (14.3) 6150 <0.001 <0.001 <0.001 A. Solhpour et al. / Atherosclerosis 207 (2009)

5 204 A. Solhpour et al. / Atherosclerosis 207 (2009) Fig. 2. Mean values of LDL-C and Apo B in different age groups in three cultures. Panel A. LDL-C male; Panel B. ApoB male; Panel C. LDL-C female; Panel D. ApoB female levels of HDL-C being the lowest of the three cultures, whereas values of apoa-i are the highest ApoB/apoA-I, TC/HDL-C, non-hdl-c/hdl ratio In both males and females, the apob/apoa-i ratio is highest in Swedes. The apob/apoa-i ratio is indistinguishable between Iranian and American males but higher in Iranian females than American females. The TC/HDL-C and non-hdl-c/hdl-c ratios are also highest in the Swedes but are significantly higher in Americans than Iranians Variance with age The differences with age of the major atherogenic lipoprotein indices are displayed in Figs Although age did not influence the relative ranking of the parameters amongst the three cultures, the mean values did vary over time. It is noteworthy that the effects of age appeared to be different in females than in males. Thus, as illustrated in Fig. 1(Panels A and B) and Fig. 2(Panels A and B) in males in all three cultures, TC, TG, LDL-C and apob levels increase substantially between the third and fifth decade, but then tend to stabilize thereafter. By contrast, in females in all three cultures, all four parameters tend to steadily increase from the third to the seventh decade (Panels 1C and 1D, 2C and 2D). HDL-C plateaus after the third decade in males (Panel 3A) whereas levels continue to climb steadily in females (Panel 3C). As illustrated in Panel 3B, levels of apoa-i are variable over the age range in males but there does not appear to be any systematic trend, whereas levels of apoa-i appear to be higher in older compared to younger females (Panel 3D). The apob/apoa-i ratio over time differs in the two genders. In males (Fig. 4, Panel 4A) it increases until the fifth decade and then changes little, whereas in females (Panel 4B) it becomes progressively greater as age increases. The same trends are seen in the TC/HDL-C ratios in males and females (Panels 4C and 4D). FBS was significantly higher in both genders in Americans compared to Swedes and significantly higher in Swedes compared to Iranians (Table 1). In females, BMI was highest in Iranians, intermediate in Americans and lowest in Swedes (Table 1). In males, BMI was highest in Americans, intermediate in Swedes and lowest in Iranians. There was no clear consistent relation between either plasma triglycerides or blood sugar and BMI. 4. Discussion INTERHEART demonstrated that the same modifiable factors interact to increase the risk of vascular disease in all the major population groups of the world [1]. This suggests the responses to injury in our arteries are much the same amongst all the major peoples of the world. On the other hand, the expression of these

6 A. Solhpour et al. / Atherosclerosis 207 (2009) Fig. 3. Mean values of HDL-C and ApoA-I in different age groups in three cultures. Panel A. HDL-C male; Panel B. ApoA-I male; Panel C. HDL-C female; Panel D. ApoA-I female risk factors amongst the peoples of the world does differ substantially. Because the plasma lipoproteins were, by far, the most important of these risk factors accounting for just over half of the population-attributable risk in INTERHEART [1], we have compared their levels in three different cultures: Sweden, Iran, and the United States. Our data demonstrate that the plasma levels of the pro- and anti-atherogenic lipoproteins differ in different cultures, they differ between the two genders, they differ over our lifespan, and finally, the major markers of the plasma lipoproteins lipids and apolipoproteins differ in how clearly they reflect these differences Limitations of study design There are important limitations in the design of this study. This study was not designed to test compare the clinical predictive powers of apolipoproteins to lipoprotein lipids. Many other studies have done so and the balance of the evidence in our view favours the apolipoproteins over the cholesterol indices [2]. In particular, INTERHEART demonstrated the predictive superiority of the apob/apoa-i ratio over any of the cholesterol ratios in all of the major peoples of the world [2]. Rather our purpose was to examine whether the relation of the major apolipoproteins to the major lipids they transport differed amongst the cultures of the world. The overall risk of vascular disease depends upon the overall risk profile. Unfortunately, our study is cross-sectional and incomplete with regard to data on all the major risk factors in all three cultures. As well inclusion/exclusion criteria are not precisely the same. That said, the sample sizes are large and the age range extensive. Comparable data are not available to determine the prevalence of vascular disease amongst the three cultures. Nevertheless, other reports demonstrate that the incidence of vascular disease appears to be comparable in Swedes and Americans [13]. Similar data are not available for Iranians but the results of surveys in Iran suggest that the prevalence of vascular disease is at least as high as in Americans and, even more interesting, there is less of a gap in risk between males and females than is the case in Western cultures [14,15]. From a methodological standpoint, the apolipoprotein measurements, which were all standardized, are superior to the conventional lipid determinations, which were obtained with a variety of methods. Accordingly, we have compared the results obtained in Iranians and Swedes to those from other studies that were done contemporaneously (see Tables 1 4 on line publication) [7,16 19]. These detailed comparisons by age and gender demonstrate only small differences from the results we have obtained Major lipoprotein findings amongst the three cultures Swedes have the highest levels of TC, LDL-C and apob but the lowest plasma triglycerides. They also have the highest levels of HDL-C, but their levels of apoa-i are very similar to Americans and the Iranians. By contrast, Iranians had the highest plasma triglycerides, LDL-C lower than Americans, but an apob that was higher than the Americans, with the lowest HDL-C but equivalent, or

7 206 A. Solhpour et al. / Atherosclerosis 207 (2009) Fig. 4. Mean values of ApoB/ApoA-I and TC/HDL-C ratios in different age groups in three cultures. Panel A. ApoB/ApoA-I ratio male; Panel B. ApoB/ApoA-I ratio female; Panel C. TC/HDL-C ratio male; Panel D. TC/HCL C ratio female. perhaps even higher, levels of apoa-i. The discordance between triglyceride and apob in the Swedes suggests that increased secretion of apob due to increased flux of fatty acids to the liver may not be the primary mechanism responsible for their high plasma apob [20]. Rather, reduced clearance of LDL, or perhaps increased secretion due to a primary cholesterol drive, are more likely to be the pathophysiological mechanisms responsible for the higher LDL levels in Swedes compared to the other two groups [20]. HDL-C levels were highest in both male and female Swedes. That HDL-C levels are higher in cultures with higher levels of cholesterol has been previously documented [21] and accounts for the fact that the inverse relation of HDL-C to cardiovascular risk is stronger within cultures than between them. We, and others, have shown that core lipid exchanges differentially affect HDL-C and apoa-i with increasing plasma triglyceride producing progressive and substantial decreases in HDL-C but relatively little change in apoa-i [22]. The discordance observed in the Swedes may reflect the obverse of this relationship. Cholesterol ester transfer protein mediated core lipid exchanges appear likely to underlie the discordance between the levels of LDL-C and apob and HDL-C and apoa-i that are so prominent in the Iranians. The combination of high triglyceride with high apob suggests that increased hepatic fatty acid flux may be an important driver of the increased apob secretion in the Iranians. The increased numbers of triglyceride-rich VLDL result in relative enrichment of LDL and HDL in triglyceride and depletion in cholesterol ester and therefore higher apob and apoa-i than LDL-C and HDL-C, respectively [2]. The precursors of atherogenic dyslipoproteinemia in the developing countries abdominal obesity and diabetes are precisely the drivers that favour the discordance between LDL-C and apob and between HDL-C and apoa-i [23]. The differences between the genders are of considerable interest. As anticipated, levels of HDL, whether expressed as HDL-C or apoa-i, are greater in females than in males in all three cultures. But, the relative order differs with HDL-C being highest in Swedish females and lowest in Iranian females with much smaller differences for apoa-i. The differences in levels of the pro- and anti-atherogenic lipoproteins over time are of particular interest. Levels of triglyceride, LDL-C, and apob in men after the third decade appear to plateau, whereas triglyceride, LDL-C and apob continue to increase in women from all three cultures. To the best of our knowledge, this discordance has not previously been highlighted and the pathophysiological bases remain to be determined. The progressive increases in the atherogenic lipoproteins might relate to greater relative decreases in age-related activity of the LDL pathway or relatively greater gain of upper body adipose tissue mass in women. Our observations that the lipoprotein-related cardiovascular risk factors have a different temporal course in females than in males accords with the known temporal difference in the incidence of clinical coronary events between the genders and sup-

8 A. Solhpour et al. / Atherosclerosis 207 (2009) ports the need for gender specific strategies for the prevention of cardiovascular disease. In summary, these data demonstrate similar associations of the pro-atherogenic and anti-atherogenic lipoproteins between the genders and over time amongst these three different cultures. They also demonstrate important discordances between the information from the lipoprotein lipids and the major apolipoproteins and that accurate and complete characterization of the lipoprotein status requires measurement of both [24]. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi: /j.atherosclerosis References [1] Yusuf S, Hawken S, Ounpuu S, et al. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study. Lancet 2004;364: [2] Barter PJ, Ballantyne CM, Carmena R, et al. ApoB versus cholesterol in estimating cardiovascular risk and in guiding therapy: report of the thirty-person/tencountry panel. J Intern Med 2006;259: [3] Marcovina SM, Gaur VP, Albers JJ. Biological variability of cholesterol, triglyceride, low- and high-density lipoprotein cholesterol, lipoprotein (a), and apolipoproteins A-I and B. Clin Chem 1994;40: [4] McQueen MJ, Hawken S, Wang X, et al. The relative importance of lipids, lipoproteins, and apolipoproteins as risk markers associated with myocardial infarction in 52 countries. Lancet 2008;372: [5] Walldius G, Jungner I, Holme I, Aastveit AH, Kolar W, Steiner E. High apolipoprotein B, low apolipoprotein A-1, and improvement in the prediction of fatal myocardial infarction (AMORIS study): a prospective study. Lancet 2001;358: [6] Karthikeyan G, Koon KT, Islam S, et al. Lipid profile, plasma apolipoproteins, and risk of a first myocardial infarction among Asians. J Am Coll Cardiol 2009;53: [7] Sarrafzadegan N, Sadri G, Malek AH, et al. Isfahan Healthy Heart Programme: a comprehensive integrated community-based programme for cardiovascular disease prevention and control. Design, methods and initial experience. Acta Cardiol 2003;58: [8] Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18: [9] Riepponen P, Marniemi J, Rautaoja T. Immunoturbidimetric determination of apolipoproteins A-1 and B in serum. Scand J Clin Lab Invest 1987;47: [10] Accessed May 30, [11] Harrell FE. Regression Modeling Strategies with Applications to Linear Models. Logistic Regression and Survival Analysis. New York: Springer; p [12] Jungner I, Marcovina SM, Walldius G, Holme I, Kolar W, Steiner E. Apolipoprotein B and A-I values in 147,576 Swedish males and females, standardized according to the World Health Organization-International Federation of Clinical Chemistry First International Reference Materials. Clin Chem 1998;44: [13] Levy D, Kannel WB. Searching for answers to ethnic disparities in cardiovascular risk. Lancet 2000;356: [14] Sarrafzadegan N, Sayed-Tabatabaei FA, Bashardoost N, et al. The prevalence of coronary artery disease in an urban population in Isfahan. Iran Acta Cardiol 1999;54: [15] Hadaegh F, Harati H, Ghanbarian A, Azizi F. Prevalence of coronary heart disease among Tehran adults: Tehran Lipid and Glucose Study. East Mediterr Health J 2009;15: [16] Ghoddusi K, Ameli J, Kachuee H, Pourfarziani V, Saadat A, Karami Q. Association of diabetes meelitus and dyslipidaemias in the Tehran population. East Mediterr Health J 2008;14: [17] Wilhelmsen L, Johansson S, Rosengren A, Wallin I, Dotevall A, Lappas G. Risk factors for cardiovascular disease during the period in Goteborg. Sweden. The GOT-MONICA Project. J Intern Med 1997;242: [18] Nilsson G, Ohrvik J, Lonnberg I, Hedberg P. Ten-year survival in 75-year-old men and women: predictive ability of total cholesterol, HDL-C and LDL-C. Curr Gerontol Geriatr Res 2009,1 7 (article ID ). [19] Berg CM, Lissner L, Aires N, et al. Trends in blood lipid levels, blood pressure, alcohol and smoking habits from 1985 to 2002: results from INTERGENE and GOT-MONICA. Eur J Cardiovasc Prev Rehabil 2005;12: [20] Sniderman AD, Zhang XJ, Cianflone K. Governance of the concentration of plasma LDL: a re-evaluation of the LDL receptor paradigm. Atherosclerosis 2000;148: [21] Knuiman JT, West CE, Katan MB, Hautvast JG. Total cholesterol and high-density lipoprotein cholesterol levels in populations differing in fat and carbohydrate intake. Arteriosclerosis 1987;7: [22] Tremblay AJ, Sniderman AD, Gagne C, Bergeron J, Couture P. Differential impact of plasma triglycerides on HDL-cholesterol and HDL-apo A-1 in a large cohort. Clin Biochem 2007;40:25 9. [23] Sarrafzadegan N, Kelishadi R, Baghaei A, et al. Metabolic syndrome: An emerging public health problem in Iranian women: Isfahan Healthy Heart Program. Int J Cardiol 2008:90 6. [24] de Graaf J, Couture P, Sniderman A. A diagnostic algorithm for the atherogenic apolipoprotein B dyslipoproteinemias. Nature Clin Endocrinol Metab 2008;4: