Lipoprotein subfractions and dietary intake of n23 fatty acid: the Genetics of Coronary Artery Disease in Alaska Natives study1–3 Giovanni Annuzzi, Angela A Rivellese, Hong Wang, Lidia Patti, Olga Vaccaro, Gabriele Riccardi, Sven OE Ebbesson, Anthony G Comuzzie, Jason G Umans, and Barbara V Howard ABSTRACT Background: Few studies have compared lipoprotein composition with dietary intake. Objective: The lipoprotein subfraction profile was evaluated in relation to diet in Alaska Eskimos at high cardiovascular risk but with a low frequency of hyperlipidemia and high intake of n23 (omega3) fatty acids. Design: A population-based sample (n = 1214) from the Norton Sound Region of Alaska underwent a physical examination and blood sampling. Analyses were from 977 individuals who did not have diabetes or use lipid-lowering medications and had complete dietary information (food-frequency questionnaire) and a lipoprotein subfraction profile (nuclear magnetic resonance spectroscopy). Results: After adjustment for age, BMI, total energy intake, and percentage of energy from fat, the intake of n23 fatty acids was significantly associated with fewer large VLDLs (P = 0.022 in women, P = 0.064 in men), a smaller VLDL size (P = 0.018 and P = 0.036), more large HDLs (P = 0.179 and P = 0.021), and a larger HDL size (P = 0.004 and P = 0.001). After adjustment for carbohydrate and sugar intakes, large VLDLs (P = 0.042 and 0.018) and VLDL size (P = 0.011 and 0.025) remained negatively associated with n23 fatty acid intake in women and men, and large HDLs (P = 0.067 and 0.005) and HDL size (P = 0.001 in both) remained positively associated with n23 fatty acid intake in women and men. In addition, large LDLs (P = 0.040 and P = 0.025) were positively associated in both sexes, and LDL size (P = 0.006) showed a positive association in women. There were no significant relations with total LDL particles in either model. Conclusions: Dietary n23 fatty acids, independent of the reciprocal changes in carbohydrate and sugar intakes, are associated with an overall favorable lipoprotein profile in terms of cardiovascular risk. Because there are no relations with total LDL particles, the benefit may be related to cardiovascular processes other than atherosclerosis. Am J Clin Nutr 2012;95:1315–22. INTRODUCTION
The role of diet in cardiovascular disease (CVD)4 is complex. Dietary fats have been shown to influence lipoprotein concentrations. Although increases in the intake of saturated fat raise the concentration of LDL cholesterol, increases in the intake of n26 polyunsaturated fat lower LDL cholesterol, and increases in the intake of n23 polyunsaturated fatty acid lower the concentration of triglycerides (1–3). A cardioprotective effect of n23 fatty acids has been shown in epidemiologic studies and clinical trials (4). Generally, changes
in fat intakes are associated with reciprocal changes in carbohydrate intakes, which may influence lipoprotein concentrations, with the most consistent effect being an increase in triglycerides and VLDLs with an increased intake of simple carbohydrates (5). In addition to plasma lipoprotein concentrations, changes in the distribution of particles within each lipoprotein class in response to dietary intakes may be relevant to the development of CVD because the atherosclerotic process may be accelerated by variations in the composition of these lipoproteins (6, 7). Proton nuclear magnetic resonance (NMR) spectroscopy provides information on the full spectrum of lipoprotein subfraction distribution and size and can be used in large population studies. However, to our knowledge, no population-based comprehensive data have been collected on the relations between dietary components and lipoprotein subfractions as determined by NMR. The Genetics of Coronary Artery Disease in Alaska Natives (GOCADAN) study is a population-based study of CVD in Alaska Eskimos that provides an opportunity to explore associations between dietary components and lipoprotein composition in a population characterized by a rapidly changing lifestyle and, thus, a wide range of diets. The GOCADAN population 1 From the Department of Clinical and Experimental Medicine, Federico II University, Naples, Italy (GA, AAR, LP, OV, and GR); the Department of Field Studies, MedStar Health Research Institute, Hyattsville, MD (HW, JGU, and BVH); the Norton Sound Health Corporation, Nome, Alaska (SOEE); the Department of Genetics, Texas Biomedical Research Institute, San Antonio, TX (AGC); and the Georgetown-Howard Universities Center for Clinical and Translational Science, Washington, DC (BVH, HW, and JGU). 2 Supported by the National Heart, Lung and Blood Institute (grants 5U01HL064244, 5U01HL08245, and 5U01HL082490). Biostatistical support for this project has been funded in whole or in part with federal funds (grant UL1RR031975) from the National Center for Research Resources, National Institutes of Health, through the Clinical and Translational Science Awards Program, which is a trademark of the US Department of Health and Human Services and part of the Roadmap Initiative “Re-Engineering the Clinical Research Enterprise.” 3 Address correspondence to BV Howard, MedStar Health Research Institute, 6525 Belcrest Road, Suite 700, Hyattsville, MD 20782. E-mail:
[email protected]. 4 Abbreviations used: CVD, cardiovascular disease; FFQ, food-frequency questionnaire; GOCADAN, Genetics of Coronary Artery Disease in Alaska Natives; IMT, intima-media thickness; NMR, nuclear magnetic resonance. Received July 25, 2011. Accepted for publication March 8, 2012. First published online May 9, 2012; doi: 10.3945/ajcn.111.023887.
Am J Clin Nutr 2012;95:1315–22. Printed in USA. Ó 2012 American Society for Nutrition
1315
1316
ANNUZZI ET AL
varies widely in intakes of fatty acids as well as of carbohydrates and sugars. Moreover, rates of CVD in Alaska Eskimos are high (8) despite high HDL cholesterol and normal LDL cholesterol and triglyceride concentrations (9). A previous study of dietary patterns in this population showed that lower triglycerides and higher HDL cholesterol were associated with a traditional dietary pattern, whereas a diet high in purchased beverages and sweets was associated with higher LDL cholesterol (10). Our analysis also showed that the intake of n23 fatty acid was negatively associated with carotid intima-media thickness (IMT) but not with the presence or extent of plaque as measured by using ultrasound (11). Another report suggested that relations with IMT vary by ethnicity (12). The variability in diets makes this population useful for exploring relations between dietary components and lipoproteins, and the availability of NMR allows the analysis of relations between dietary components and lipoprotein subclasses and thus provides an additional understanding of the role of diet in the modulation of cardiovascular risk. In this article, relations between n23 fatty acid intake and lipoprotein particle concentrations and size are explored, and potential confounding by carbohydrate intake is examined in Alaska Eskimo participants of the GOCADAN study.
SUBJECTS AND METHODS
Study population Details of the study protocol have been published (13, 14). The GOCADAN population includes 1214 family members, aged 17– 91 y, who are residents of 8 villages and the town of Nome in the Norton Sound Region of Alaska. Participants were recruited in 2000–2004. Except for one village, 74% of all age-eligible residents participated. Each participant underwent a physical examination, personal interview, collection of biological specimens, and diagnostic tests. Permission to conduct the study was granted by the Norton Sound Health Corporation and the institutional review board of each institution. Written informed consent was obtained from all participants. For the current analysis, only participants with complete data on diet and lipoprotein particle measurements were included (n = 1059). Of the 1059 participants, individuals with diabetes as defined by 1998 WHO criteria (15) (n = 37) or who reported taking hypolipidemic agents (n = 62) were excluded. After these exclusions, 977 subjects were included in the final data set. Measurements Participants were seen in the fasting state. Anthropometric measurements, including height, weight, and waist circumference, were performed according to standard procedures (13). Weight was measured to the nearest one-tenth of a pound. Waist circumference was measured with an anthropometric tape applied at the level of the umbilicus with the subject supine and was approximated to the nearest one-quarter of an inch. Samples of whole blood, plasma, serum, and urine were collected from each participant and stored at –80°C until used. All laboratory methods have been published (13). Plasma lipid concentrations were analyzed by using a conventional enzymatic chemistry analyzer (Vitros 950; Ortho-Clinical Diagnostics)
with a dry multilayered analytic element coated on a polyester support (16, 17). LDL cholesterol was calculated by using Friedewald’s formula (18). Lipoprotein subfraction profile A detailed lipoprotein subclassification was performed on plasma EDTA isolated by centrifugation (3000 rpm for 10 min at 4°C) and stored at –80°C for the NMR spectroscopy (19) with the use of a rapid, automated, commercially available assay (LipoScience Inc). Details of the NMR methodology have been published (19). Briefly, the NMR technique uses the characteristic spectral signal broadcast by lipoprotein subfractions of different sizes as the basis for the quantification. The amplitude of the emitted signal is proportional to the number of terminal methyl groups of the 4 lipids (phospholipid, cholesterol, cholesterol ester, and triglyceride) contained within the particle of that size range and provides a direct measure of the particles present. The measured amplitudes of these subclass signals are directly proportional to the number of particles that emit the signal. Therefore, VLDL, LDL, and HDL subclasses can be estimated from the intensity of their NMR signals with the use of conversion factors. Weighted VLDL, LDL, and HDL particle sizes (nm diameter) were computed as the sum of the diameter of each subclass and multiplied by its relative mass percentage as estimated from the amplitude of its NMR signal. The following spectra resulted: large (V5 + V6, 61.0–220.0 nm), intermediate (V3 + V4, 36.0–60.9 nm), and small VLDL (V1 + V2, 27.0–35.9 nm); large (L3, 21.3–22.7 nm), intermediate (L2, 19.8–21.2 nm), and small LDL (L1, 18.3–19.7 nm); and large (H4 + H5, 8.8–18.3 nm), intermediate (H3, 8.2–8.7 nm), and small HDL (H1 + H2, 7.3–8.1 nm). Data on intermediate-density lipoprotein particles (22.7–27 nm) were not reported in the current analysis. For all analyses, intermediate and small LDL particles were grouped together and considered small particles, as has been done in other studies (20–22), and intermediate HDL were combined with small HDL particles because of their small contribution to total HDL. Dietary evaluations Dietary habits were evaluated by using a food-frequency questionnaire (FFQ), which measured consumption during the previous year, on the basis of an FFQ that was previously developed for use in the region and validated with 24-h recalls (23). The original FFQ was modified to include all major traditional and key foods (a total of 97 foods) that are available in small village stores and more details about specific fats and food preparation. The diet interviews were not conducted during any specific time of the year. The 20- to 30-min diet interview was conducted at the survey site in each community as part of the 2-h clinical interview. Interviewers used 3-dimensional food models to determine the amount, frequency, and seasonality of each food consumed. Quality-assurance procedures were adhered to during data collection, entry, and analysis. When the portion size was omitted, one standard portion was assumed; for omissions of season (7 records), all year was assumed, except for the categories bird and bird eggs, which, on the basis of reports of the other participants, were assumed to be consumed during the 10-wk spring season. Seal and salmon were the 2 main dietary
1317
DIETARY n–3 FATTY ACIDS AND LIPOPROTEIN SUBFRACTIONS
sources of n23 fatty acids. Nutrient calculations were created with Nutrition Data System for Research (database version 4.06_34; University of Minnesota Nutrition Coordinating Center). Statistical analysis Data are presented as means 6 SDs, medians (IQRs), or geometric means (95% CIs) as specified. Sex differences were evaluated by using Student’s t test for variables that were normally distributed and by Wilcoxon’s rank-sum test for variables not normally distributed. Relations between dietary components and lipoprotein subfractions (expressed as concentrations and sizes) were evaluated by using the quartile of intake for each major macronutrient. Comparisons among groups were performed by using ANOVA with adjustment for age and BMI. Because of the observed relations between n23 intakes and intakes of total energy and percentage of energy from fat, these measures were included in all models; models were run with and without adjustment for intakes of carbohydrate and simple sugars. Spearman’s partial correlation coefficients between n23 fatty acids and lipoproteins were calculated; age, BMI, total energy intake, and intakes of total fat, carbohydrates, and simple sugars were controlled for. A natural log transformation was applied to variables that were highly skewed. Back-transformed mean values are listed in the tables, where indicated. However, the transformed variables were used in the statistical analyses. All P values were 2-tailed, and P , 0.05 was considered statistically significant. Data were analyzed with SAS software (version 9.1; SAS Institute). RESULTS
The main characteristics of the participants by sex are shown in Table 1. Women had significantly higher BMI and insulin concentrations as well as lower blood pressure and higher HDL cholesterol. As previously reported (2), the lipoprotein profile
differed by sex (Table 2). Compared with men, women had fewer total VLDL particles, fewer total LDL particles (because of markedly fewer small particles despite higher concentrations of the larger ones), and higher concentrations of total and large HDL. The main dietary components consumed, as determined by the FFQ, are shown in Table 2. In both sexes, the diet was rich in total fat (mean: 37%) and simple sugars (mean: 27%). As expected, the main sex difference in diet was a higher total energy intake in men. The intake of n23 fatty acids [median: 2.6 g/d (0.8% of energy) and 3.0 g/d (0.7% of energy) for women and men, respectively] was higher than generally observed in Western populations (24) but may be similar to that of some Japanese groups (25). Relations between dietary habits and lipoprotein profiles were evaluated by comparing the concentrations and sizes of lipoprotein fractions in women and men stratified by quartiles of nutrient intakes. In a multivariate model adjusted for age and BMI, significant trends were shown in key lipoprotein measures with an increasing total energy intake as follows: fewer total VLDLs (mean 6 SD across increasing quartiles: 64.6 6 28.7, 66.1 6 33.3, 59.4 6 27.9, and 57.9 6 26.0 nmol/L; P-trend = 0.043) and small VLDL (45.5 6 19.5, 46.5 6 22.1, 42.3 6 18.8, and 40.0 6 17.9 nmol/L; P-trend = 0.021) in women and more HDL particles (26.2 6 5.8, 27.4 6 5.7, 26.8 6 5.3, and 28.1 6 6.0 nmol/L; P-trend = 0.043) in men. Significant trends were also shown with increasing total fat intakes as follows: fewer large VLDLs (0.67 6 6.20, 0.55 6 6.44, 0.39 6 6.36, and 0.43 6 7.47 nmol/L; P-trend = 0.014) and greater HDL size (9.11 6 0.47, 9.23 6 0.50, 9.24 6 0.46, and 9.32 6 0.47 nmol/L; Ptrend = 0.014) in women and fewer large VLDLs (0.85 6 5.17, 0.56 6 6.89, 0.48 6 7.48, and 0.47 6 7.46 nmol/L; P-trend = 0.046) in men. Therefore, all subsequent analyses were adjusted for total energy and total fat intake. In women (Table 3; see supplemental figure under “Supplemental data” in the online issue), a higher n23 dietary intake
TABLE 1 Clinical and biochemical characteristics of participants by sex
Age (y) BMI (kg/m2) Waist circumference (cm) Plasma glucose (mmol/L) Plasma insulin (pmol/L) Plasma triglycerides (mmol/L) Plasma cholesterol (mmol/L) LDL cholesterol (mmol/L) HDL cholesterol (mmol/L) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Current smoker (%) Current drinker (%)4 Hypertension (%) Use of antihypertensive medications (%) 1
Women (n = 559)
Men (n = 418)
P1
40.7 6 15.22 28.1 6 6.0 87.4 6 13.6 5.1 6 0.5 62.1 (59.3, 65.0)3 1.2 (1.2, 1.3) 5.3 6 1.1 3.0 6 0.9 1.7 6 0.5 116.2 6 15.1 74.1 6 8.8 59.2 67.4 13.4 9.1
41.2 6 14.8 26.4 6 4.9 87.1 6 12.0 5.1 6 0.6 49.8 (46.9, 53.0) 1.2 (1.2, 1.3) 5.1 6 1.0 3.0 6 0.9 1.4 6 0.5 121.2 6 12.6 78.5 6 9.3 62.0 75.4 20.6 7.4
0.624 ,0.001 0.770 0.226 ,0.001 0.840 0.027 0.278 ,0.001 ,0.0001 ,0.0001 0.385 0.007 0.003 0.341
Differences were evaluated by using Student’s t test for variables that were normally distributed and Wilcoxon’s rank-sum test for variables that were not normally distributed. 2 Mean 6 SD (all such values). 3 Geometric mean; 95% CI in parentheses (all such values). 4 Defined as any alcohol intake in the past year.
1318
ANNUZZI ET AL TABLE 2 Lipoprotein particle distribution and size and dietary intake evaluated by using a food-frequency questionnaire
Lipoproteins Total VLDL particles (nmol/L) Large VLDLs and chylomicrons (nmol/L) Medium VLDLs (nmol/L) Small VLDLs (nmol/L) VLDL size (nm) Total LDL particles (nmol/L) Large LDLs (nmol/L) Small LDLs (nmol/L) LDL size (nm) Total HDL particles (lmol/L) Large HDLs (lmol/L) Small HDLs (lmol/L) Small and medium HDLs (lmol/L) HDL size (nm) Dietary intake Total energy (kcal/d) Total fat (percentage of total energy) Saturated fat (percentage of total energy) Monounsaturated fat (percentage of total energy) n26 PUFA (percentage of total energy) n23 PUFA (percentage of total energy) Protein (percentage of total energy) Carbohydrate (percentage of total energy) Simple sugars (percentage of total energy) Cholesterol (mg/d)
P1
Women
Men
58.3 6 26.82 0.9 (0.1, 2.4)3 14.6 (6.1, 22.2) 41.2 6 18.6 45.3 6 8.5 1025 6 292 554 6 184 450 6 295 21.7 6 0.6 29.6 6 6.7 7.2 6 3.8 20.4 6 5.3 22.4 6 6.0 9.2 6 0.5
62.0 6 29.2 1.0 (0.2, 2.6) 14.7 (6.8, 24.1) 43.6 6 19.7 46.1 6 9.8 1118 6 349 493 6 172 605 6 349 21.4 6 0.6 27.2 6 5.7 5.1 6 3.7 20.3 6 4.7 22.1 6 5.2 9.0 6 0.5
0.039 0.187 0.279 0.051 0.195 ,0.001 ,0.001 ,0.001 ,0.001 ,0.001 ,0.001 0.710 0.3179 ,0.001
6 6 6 6 6 6 6 6 6 6
,0.001 0.911 0.763 0.712 0.325 0.133 0.094 0.471 0.204 ,0.001
2874 37.2 12.7 14.2 6.3 0.8 14.5 49.2 27.9 466
6 6 6 6 6 6 6 6 6 6
1434 8.7 3.6 3.6 2.4 0.8 4.5 11.0 12.1 293
3514 37.2 12.8 14.2 6.0 0.7 14.9 48.7 26.9 629
1575 8.6 3.5 3.6 2.0 0.9 4.0 10.8 12.1 375
1
Differences were evaluated by Student’s t test for variables that were normally distributed and Wilcoxon’s rank-sum test for variables not normally distributed. 2 Mean 6 SD (all such values). 3 Geometric mean; 95% CI in parentheses (all such values).
was associated with significantly fewer large VLDL particles, a smaller VLDL size, fewer small and medium HDL particles, and a larger HDL size. In men, a higher intake of n23 fatty acids was associated with a trend for fewer large VLDLs (P = 0.064), significantly more small VLDL particles, and a significantly higher large HDL particle concentration and larger HDL size. More large LDLs (P = 0.051) and a larger LDL size (P = 0.020) were also associated with a higher n23 intake in men. Intakes of n23 fatty acids and carbohydrates and sugars were inversely related. Higher intakes of n23 (from the first to fourth quartiles) were associated with a lower percentage of intake of carbohydrates and simple sugars in women (carbohydrates: 52.6– 44.3%, P , 0.0001; sugars: 30.4–24.5%, P , 0.0001) and in men (carbohydrates: 51.7–45.2%, P , 0.0001; sugars: 29.5–24.4%, P = 0.002). When relations between carbohydrate intakes and lipoproteins were examined, higher intakes of total carbohydrates were associated with higher concentrations of large VLDLs and chylomicrons in both sexes, and higher sugar intakes were associated with lower concentrations of large HDLs in women (P = 0.025) and higher concentrations of small HDLs in men (P = 0.046). We evaluated whether relations between n23 fatty acids and lipoproteins were independent of carbohydrate and sugar intakes (Table 4). After additional adjustment for carbohydrates and simple sugars, all relations except small VLDLs in men remained significant; in addition, n23 fatty acid intakes were correlated with higher large LDL concentrations in women and men and with a larger LDL size in men. When analyses
were run that included smoking and alcohol intake in the model, results were unchanged. DISCUSSION
To our knowledge, this study was the first to examine relations between the dietary intake of n23 fatty acid and the lipoprotein profile as evaluated by using NMR spectroscopy. This analysis was conducted in a population with a rapidly changing lifestyle that has led to wide ranges in n23 fatty acid intakes. A higher intake of n23 fatty acid was associated with lower concentrations of large VLDLs, a smaller average VLDL size, a higher large HDL concentration, and a larger HDL size in both sexes. Large LDL concentration and LDL size were higher with an increasing intake of n23 only in men and small and medium HDLs were lower only in women. These relations were independent of age, obesity, total energy intake, and total fat intake. Although carbohydrate and sugar intakes also were related to VLDLs, the differences in lipoprotein variables with higher n23 intakes were independent of changes in carbohydrate and sugar intakes, and relations with large LDLs became significant in women. A large body of evidence from observational studies and clinical trials has clarified that higher intakes of foods or dietary supplements that contain n23 fatty acids lead to lower plasma triglyceride concentrations and higher HDL cholesterol (3). The current analysis of lipoprotein subclass distribution provides
1319
DIETARY n–3 FATTY ACIDS AND LIPOPROTEIN SUBFRACTIONS TABLE 3 Lipoprotein particle concentration and size by quartile of n23 fatty acid intake in women and men
Women n Range of n23 fatty acid intake (g/d) Total VLDL particles (nmol/L) Large VLDLs and chylomicrons (nmol/L)3 Medium VLDLs (nmol/L)3 Small VLDLs (nmol/L) VLDL size (nm) Total LDL particles (nmol/L) Large LDLs (nmol/L) Small LDLs (nmol/L) LDL size (nm) Total HDL particles (lmol/L) Large HDLs (lmol/L) Small HDLs (lmol/L) Small and medium HDLs (lmol/L) HDL size (nm) LDL cholesterol (mg/dL)4 HDL cholesterol (mg/dL)4 Triglycerides (mg/dL)4 Men n Range of n23 fatty acid intake (g/d) Total VLDL particles (nmol/L) Large VLDLs and chylomicrons (nmol/L)3 Medium VLDLs (nmol/L)3 Small VLDLs (nmol/L) VLDL size (nm) Total LDL particles (nmol/L) Large LDLs (nmol/L) Small LDLs (nmol/L) LDL size (nm) Total HDL particles (lmol/L) Large HDLs (lmol/L) Small HDLs (lmol/L) Small and medium HDLs (lmol/L) HDL size (nm) LDL cholesterol (mg/dL)4 HDL cholesterol (mg/dL)4 Triglycerides (mg/dL)4 1 2 3 4
First quartile
Second quartile
Third quartile
Fourth quartile
141 0.02–0.65 59.5 6 26.82 0.63 6 6.03 11.10 6 2.54 42.3 6 17.9 46.3 6 8.4 1031 6 336 526 6 186 479 6 329 21.64 6 0.63 30.58 6 6.38 6.53 6 3.83 21.17 6 5.34 24.05 6 5.69 9.08 6 0.46 113.74 6 36.93 62.45 6 17.41 114.62 6 1.68
138 0.66–1.54 56.1 6 23.8 0.53 6 5.83 10.41 6 3.26 39.2 6 17.1 46.4 6 8.9 1004 6 285 537 6 185 448 6 288 21.70 6 0.63 28.58 6 6.52 7.17 6 3.66 19.59 6 4.95 21.41 6 5.79 9.22 6 0.48 108.41 6 33.30 62.57 6 17.97 110.63 6 1.56
139 1.55–3.21 60.4 6 29.6 0.52 6 7.47 10.40 6 3.26 42.4 6 19.9 45.8 6 8.9 1069 6 284 588 6 187 458 6 294 21.75 6 0.61 29.56 6 7.26 7.26 6 3.99 20.33 6 5.60 22.30 6 6.79 9.24 6 0.49 119.99 6 36.84 64.21 6 17.29 112.60 6 1.62
141 3.22–29.6. 57.2 6 26.9 0.35 6 7.17 9.87 6 3.25 40.8 6 19.3 42.7 6 7.2 1000 6 257 564 6 173 416 6 266 21.78 6 0.59 29.61 6 6.59 7.73 6 3.78 20.48 6 5.05 21.89 6 5.53 9.35 6 0.46 118.33 6 34.65 66.19 6 17.37 98.57 6 1.56
— — 0.137 0.022 0.164 0.121 0.018 0.389 0.363 0.232 0.260 0.300 0.179 0.549 0.044 0.004 0.0585 0.0517 0.0150
105 0.03–0.71 58.3 6 24.9 0.80 6 5.57 10.61 6 3.33 40.5 6 17.2 47.41 6 9.17 1089 6 336 447 6 172 622 6 339 21.25 6 0.57 26.52 6 5.95 4.21 6 3.12 20.28 6 4.91 22.31 6 5.02 8.79 6 0.41 108.96 6 33.37 48.90 6 14.20 116.00 6 1.69
104 0.72–1.70 63.5 6 30.7 0.66 6 6.75 10.95 6 3.39 44.6 6 19.8 46.51 6 9.22 1123 6 318 499 6 168 601 6 308 21.40 6 0.59 27.21 6 5.87 4.92 6 3.70 20.35 6 4.71 22.29 6 5.32 8.92 6 0.50 116.54 6 35.32 54.63 6 15.94 111.59 6 1.63
105 1.71–3.41 64.7 6 28.7 0.47 6 6.66 12.87 6 2.74 44.8 6 19.1 45.15 6 10.62 1106 6 355 499 6 168 590 6 381 21.43 6 0.64 27.94 6 5.38 5.88 6 4.15 20.27 6 4.36 22.06 6 5.22 9.11 6 0.57 119.49 6 34.58 59.54 6 22.01 105.59 6 1.67
104 3.44–65.3 61.5 6 32.0 0.44 6 7.96 9.10 6 3.94 44.3 6 22.4 45.26 6 9.92 1156 6 385 527 6 173 606 6 368 21.45 6 0.58 26.94 6 5.66 5.42 6 3.41 20.19 6 4.83 21.53 6 5.36 9.05 6 0.52 125.64 6 40.16 57.53 6 16.76 105.59 6 1.58
— — 0.113 0.064 0.897 0.036 0.131 0.783 0.051 0.469 0.020 0.933 0.021 0.850 0.153 0.0001 0.0008 ,0.0001 0.1228
P-trend1
Multivariate model was adjusted for age, BMI, total energy intake, and percentage of energy from fat. Unadjusted mean 6 SD (all such values). Values were back log transformed. Plasma lipid values were determined by using conventional methods.
insight into the mechanisms of these changes. The observed lower large-VLDL concentration and smaller VLDL size with higher n23 intakes may reflect enhanced catabolism of VLDL, leading to lower concentrations of circulating triglycerides. Although concentrations of smaller VLDLs were not decreased, larger LDL particles in men were observed with higher n23 intakes in the current analysis. These relations between the n23 fatty acid intake and lipoprotein particles in a general population are compatible with kinetic studies in insulin-resistant obese men with dyslipidemia that show an increased conversion of VLDL to LDL particles with fish-oil supplements (26). In turn, the accelerated catabolism increases the availability of triglycerides for exchange with the HDL compartment; this mechanism might explain our observed increasing large and decreasing
small and medium HDL concentrations and higher particle size with higher intakes of n23 fatty acids. This finding is also consistent with previous studies of n23 fatty acid supplementation showing that the cholesterol content shifts from HDL3 particles to the larger HDL2 particles (27). In the presence of enhanced catabolism of large VLDLs, reduced cholesteryl ester transfer protein activity (28) would be expected to reduce the triglyceride and cholesterol ester exchange with HDL. Because this activity results in less replacement of HDL cholesterol with triglycerides, this eventually may lead to the prevalence of large HDL particles because the relative lack of triglycerides makes them more resistant to the action of hepatic lipase (29). At the same time, the impaired cholesteryl ester transfer protein–mediated triglyceride transfer to LDL would also increase the proportion of cholesterol
1320
ANNUZZI ET AL
TABLE 4 Partial correlations between n23 fatty acid intake (g/d) and lipoprotein subfractions controlled for age, BMI, and total intakes (%) of energy, fat, carbohydrates, and sugars Correlation coefficient Women Total VLDL particles (nmol/L) Large VLDLs and chylomicrons (nmol/L) Medium VLDLs (nmol/L) Small VLDLs (nmol/L) VLDL size (nm) Total LDL particles (nmol/L) Large LDLs (nmol/L) Small LDLs (nmol/L) LDL size (nm) Total HDL particles (lmol/L) Large HDLs (lmol/L) Small HDLs (lmol/L) Small and medium HDLs (lmol/L) HDL size (nm) Men Total VLDL particles (nmol/L) Large VLDLs and chylomicrons (nmol/L) Medium VLDLs (nmol/L) Small VLDLs (nmol/L) VLDL size (nm) Total LDL particles (nmol/L) Large LDLs (nmol/L) Small LDLs (nmol/L) LDL size (nm) Total HDL particles (lmol/L) Large HDLs (lmol/L) Small HDLs (lmol/L) Small and medium HDLs (lmol/L) HDL size (nm)
P
20.054 20.087 20.045 20.059 20.108 20.018 0.087 20.040 0.057 20.074 0.078 20.049 20.115 0.162
0.205 0.042 0.287 0.163 0.011 0.672 0.040 0.348 0.179 0.081 0.067 0.246 0.007 0.001
0.023 20.116 20.005 0.053 20.111 20.036 0.111 20.110 0.136 0.009 0.137 20.011 20.084 0.214
0.648 0.018 0.915 0.280 0.025 0.473 0.025 0.026 0.006 0.851 0.005 0.828 0.090 ,0.001
in LDL and, thus, increase the prevalence of large LDLs, as was observed in men in the current study. In contrast, the differences observed with n23 fatty acid intakes in the current analysis may have resulted from effects on hepatic triglyceride and VLDL production (26); the specific physical properties of n23 fatty acids may favor the output of smaller VLDL particles. Again, this possibility could explain the presence of larger LDLs (30), but alternative mechanisms, perhaps driven by the specific effects of n23 fatty acids on phospholipid configuration, are needed to explain the higher number of large HDL particles. These possibilities cannot be evaluated further in observational analyses and await additional metabolic studies. Relations between n23 intakes and lipoprotein fractions were independent of the carbohydrate and sugar intakes that were associated with opposite trends for lipoprotein classes. In particular, after correction for carbohydrate and sugar intakes, relations between n23 and LDL subclasses were enhanced in both sexes. It is known that an increased intake of carbohydrates, particularly of simple sugars and carbohydrates with a high glycemic index, can increase concentrations of small, dense LDLs and HDLs (31), but because the adjustment enhanced the relations, our analyses suggested that the observed relations with n23 fatty acids were not due to differences in carbohydrate intakes.
In the current study, the differences in lipoprotein profiles with higher n23 fatty acid intakes were generally favorable in terms of CVD risk. Observational studies and clinical trials have established that the intake of n23 fatty acids, either from fish or dietary supplementation, is associated with lower rates of CVD (4). Clinical trial data have suggested that this benefit was mainly due to decreases in sudden death and arrhythmic events (32, 33), but some trials have also shown reductions in atherothrombotic events (34) and vascular reactivity (35). HDL particles are known to have antiinflammatory properties, protect LDL from oxidation, and improve endothelial function (36); thus, the increase in HDL particle concentrations would be associated with these beneficial effects and is consistent with reduced CVD events. Moreover, large HDL particles (HDL2) are thought to be more atheroprotective than small HDL (HDL3) (37, 38). In the Multi-Ethnic Study of Atherosclerosis cohort, NMR studies of lipoprotein particles also confirmed that large HDL particles were inversely correlated with IMT (20). Recent analyses of the Multi-Ethnic Study of Atherosclerosis study also suggested an inverse relation between large HDLs and CVD events (39). The role of VLDL in CVD is less well understood. Large VLDLs are believed to be less atherogenic, whereas small VLDLs have been shown to penetrate the vessel intima (40, 41). We observed a smaller VLDL size with a higher n23 fatty acid intake, but this effect may not increase atherosclerosis because it was not accompanied by an increase in small particles. We previously reported that the major associations between lipoprotein composition and carotid measures are with LDL in the GOCADAN population, in which higher concentrations of LDL particles were associated with increasing IMT and smaller VLDLs, but a larger LDL size was associated with plaque (22). n23 Fatty acids have been reported to modulate atherosclerosis progression by inhibiting cholesterol delivery to the arterial wall and by reducing total LDL cholesterol ester uptake and LDLselective uptake through the regulation of arterial lipoprotein lipase concentrations (42). Although the associations with IMT are consistent with a beneficial effect of n23 intake, the current associations of n23 with larger LDL size in men and smaller VLDLs in both sexes would appear to have an adverse effect. However, the LDL particle number has been shown to be a more important determinant than LDL size (6, 7, 20, 37). More studies are needed to understand the complex relations between VLDL and LDL compositions and CVD. In this population, other factors, such as high levels of inflammation from subclinical infections (43) and high smoking rates (8) may also modulate lipoprotein particle-vessel wall interactions. The strengths of this study include the large, population-based sample, the availability of dietary data obtained with an FFQ developed for use with this population, and the availability of standardized lipoprotein subfraction data. This cohort is relatively homogeneous, with a wide range of n23 fatty acid intakes. On the other hand, this study was confined to one unique ethnic group, and data are needed in other populations. Also, because the diet analysis focused on nutrients, it was not possible to determine whether unique aspects of specific foods might be responsible for some of the findings. This study was limited by its cross-sectional design and an examination of various closely related variables. To address this limitation, we narrowed our analyses and used fully adjusted models. A confirmation of the current findings awaits longitudinal and metabolic studies.
DIETARY n–3 FATTY ACIDS AND LIPOPROTEIN SUBFRACTIONS
In conclusion, lifestyles are rapidly changing in this unique population. These lifestyle changes are associated with dietary changes; persons who consume a traditional diet consume more n23 fatty acids and less carbohydrates and sugars, whereas individuals with a more Western lifestyle often consume diets that are high in carbohydrates and sugars (10). The current data suggest that this dietary transition influences lipoprotein concentrations and sizes, with the changes, on balance, being adverse in terms of cardiovascular risk. Dietary n23 fatty acids are associated with a lipoprotein profile that is, overall, favorable in terms of cardiovascular risk. Because no relations with LDL particle numbers were observed, the benefit may be related to cardiovascular processes other than atherosclerosis. These findings should be considered in the implementation of education programs to counsel individuals on healthier food choices when traditional foods are not available. We acknowledge the assistance and cooperation of the Eskimo communities of the Norton Sound region, Alaska, without whose support this study would not have been possible. We thank Rachel Schaperow, MedStar Health Research Institute, for editing the manuscript. The authors’ responsibilities were as follows—AGC, BVH, SOEE, AAR, and GR: designed the research (project conception, development of overall research plan, and study oversight); JGU and SOEE: conducted the research (hands-on conduct of the experiments and data collection), provided essential reagents, and provided essential materials; HW and OV: analyzed the data and performed statistical analyses; GA, AAR, LP, and BVH: wrote the article; and GA and BVH: had primary responsibility for the final content of the manuscript. BVH has served on the advisory boards of Merck/Schering Plough. The other authors had nothing to declare.
REFERENCES 1. Riccardi G, Giacco R, Rivellese AA. Dietary fat, insulin sensitivity and the metabolic syndrome. Clin Nutr 2004;23:447–56. 2. Rivellese AA, Maffettone A, Vessby B, Uusitupa M, Hermansen K, Berglund L, Louheranta A, Meyer BJ, Riccardi G. Effects of dietary saturated, monounsaturated and n23 fatty acids on fasting lipoproteins, LDL size and post-prandial lipid metabolism in healthy subjects. Atherosclerosis 2003;167:149–58. 3. Hartweg J, Farmer AJ, Perera R, Holman RR, Neil HA. Meta-analysis of the effects of n23 polyunsaturated fatty acids on lipoproteins and other emerging lipid cardiovascular risk markers in patients with type 2 diabetes. Diabetologia 2007;50:1593–602. 4. Wang C, Harris WS, Chung M, Lichtenstein AH, Balk EM, Kupelnick B, Jordan HS, Lau J. n23. Fatty acids from fish or fish-oil supplements, but not a-linolenic acid, benefit cardiovascular disease outcomes in primary- and secondary- prevention studies: a systematic review. Am J Clin Nutr 2006;84:5–17. 5. Sacks FM, Katan M. Randomized clinical trials on the effects of dietary fat and carbohydrate on plasma lipoproteins and cardiovascular disease. Am J Med 2002;113(suppl 9B):13S–24S. 6. El Harchaoui K, van der Steeg WA, Stroes ES, Kuivenhoven JA, Otvos JD, Wareham NJ, Hutten BA, Kastelein JJ, Khaw KT, Boekholdt SM. Value of low-density lipoprotein particle number and size as predictors of coronary artery disease in apparently healthy men and women: the EPIC-Norfolk Prospective Population Study. J Am Coll Cardiol 2007; 49:547–53. 7. Carmena R, Duriez P, Fruchart JC. Atherogenic lipoprotein particles in atherosclerosis. Circulation 2004;109(Suppl 1):III2–7. 8. Howard BV, Comuzzie A, Devereux RB, Ebbesson SO, Fabsitz RR, Howard WJ, Laston S, MacCluer JW, Silverman A, Umans JG, et al. Cardiovascular disease prevalence and its relation to risk factors in Alaska Eskimos. Nutr Metab Cardiovasc Dis 2010;20:350–8. 9. Rivellese AA, Patti L, Kaufman D, Zhu J, Annuzzi G, Vaccaro O, Ebbesson SO, Howard WJ, Riccardi G, Howard BV. Lipoprotein particle distribution and size, insulin resistance, and metabolic syndrome in Alaska Eskimos: the GOCADAN study. Atherosclerosis 2008;200: 350–8.
1321
10. Eilat-Adar S, Mete M, Nobmann ED, Xu J, Fabsitz RR, Ebbesson SO, Howard BV. Dietary patterns are linked to cardiovascular risk factors but not to inflammatory markers in Alaska Eskimos. J Nutr 2009;139: 2322–8. 11. Ebbesson SO, Roman MJ, Devereux RB, Kaufman D, Fabsitz RR, MacCluer JW, Dyke B, Laston S, Wenger CR, Comuzzie AG, et al. Consumption of omega-3 fatty acids is not associated with a reduction in carotid atherosclerosis: the Genetics of Coronary Artery Disease in Alaska Natives study. Atherosclerosis 2008;199:346–53. 12. Sekikawa A, Curb JD, Ueshima H, El-Saed A, Kadowaki T, Abbott RD, Evans RW, Rodriguez BL, Okamura T, Sutton-Tyrrell K, et al. Marine-derived n23 fatty acids and atherosclerosis in Japanese, Japanese-American, and white men: a cross-sectional study. J Am Coll Cardiol 2008;52:417–24. 13. Howard BV, Devereux RB, Cole SA, Davidson M, Dyke B, Ebbesson SO, Epstein SE, Robinson DR, Jarvis B, Kaufman DJ, et al. A genetic and epidemiologic study of cardiovascular disease in Alaska Natives (GOCADAN): design and methods. Int J Circumpolar Health 2005;64: 206–21. 14. Ebbesson SO, Laston S, Wenger CR, Dyke B, Romenesko T, Swenson M, Fabsitz RR, MacCluer JW, Devereux R, Roman M, et al. Recruitment and community interactions in the GOCADAN study. Int J Circumpolar Health 2006;65:55–64. 15. The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 1997;20:1183–97. 16. Allain CC, Poon LS, Chan CS, Richmond W, Fu PC. Enzymatic determination of total serum cholesterol. Clin Chem 1974;20:470–5. 17. Spayd RW, Bruschi B, Burdick BA, Dappen GM, Eikenberry JN, Esders TW, Figueras J, Goodhue CT, LaRossa DD, Nelson RW, et al. Multilayer film elements for clinical analysis: applications to representative chemical determinations. Clin Chem 1978;24:1343–50. 18. Friedewald WT, Levy RI, Fredrikson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18:499–502. 19. Jeyarajah EJ, Cromwell WC, Otvos JD. Lipoprotein particle analysis by nuclear magnetic resonance spectroscopy. Clin Lab Med 2006;26:847–70. 20. Mora S, Szklo M, Otvos JD, Greenland P, Psaty BM, Goff DC Jr, O’Leary DH, Saad MF, Tsai MY, Sharrett AR. LDL particle subclasses, LDL particle size, and carotid atherosclerosis in the Multi-Ethnic Study of Atherosclerosis (MESA). Atherosclerosis 2007;192:211–7. 21. Festa A, Williams K, Hanley AJ, Otvos JD, Goff DC, Wagenknecht LE, Haffner SM. Nuclear magnetic resonance lipoprotein abnormalities in prediabetic subjects in the Insulin Resistance Atherosclerosis Study. Circulation 2005;111:3465–72. 22. Masulli M, Patti L, Riccardi G, Vaccaro O, Annuzzi G, Ebbesson SO, Fabsitz RR, Howard WJ, Otvos JD, Roman MJ, et al. Relation among lipoprotein subfractions and carotid atherosclerosis in Alaskan Eskimos (from the GOCADAN Study). Am J Cardiol 2009;104:1516–21. 23. Nobmann EAD. Diet among Siberian Yup’iks of Alaska and the implications for cardiovascular disease. PhD dissertation, University of Alaska, Fairbanks, AK, 1996. 24. Nobmann ED, Ponce R, Mattil C, Devereux R, Dyke B, Ebbesson SO, Laston S, MacCluer J, Robbins D, Romenesko T, et al. Dietary intakes vary with age among Eskimo adults of Northwest Alaska in the GOCADAN study, 2000-2003. J Nutr 2005;135:856–62. 25. Iso H, Kobayashi M, Ishihara J, Sasaki S, Okada K, Kita Y, Kokubo Y, Tsugane S; JPHC Study Group. Intake of fish and n3 fatty acids and risk of coronary heart disease among Japanese: the Japan Public Health Center-Based (JPHC). Study Cohort I.Circulation. 2006;113:195–202. 26. Chan DC, Watts GF, Barrett PHR, Beilin LJ, Redgrave TG, Mori TA. Regulatory effects of HMG CoA reductase inhibitor and fish oils on apolipoprotein B-100 kinetics in insulin-resistant obese male subjects with dyslipidemia. Diabetes 2002;51:2377–86. 27. Wooten JS, Biggerstaff KD, Ben-Ezra V. Responses of LDL and HDL particle size and distribution to omega-3 fatty acid supplementation and aerobic exercise. J Appl Physiol 2009;107:794–800. 28. Pownall HJ, Brauchi D, Kilinc¸ C, Osmundsen K, Pao Q, Payton-Ross C, Gotto AM Jr, Ballantyne CM. Correlation of serum triglyceride and its reduction by omega-3 fatty acids with lipid transfer activity and the neutral lipid compositions of high-density and low-density lipoproteins. Atherosclerosis 1999;143:285–97. 29. Taskinen M-R. Diabetic dyslipidaemia: from basic research to clinical practice. Diabetologia 2003;46:733–49.
1322
ANNUZZI ET AL
30. Mori TA, Burke V, Puddey IB, Watts GF, O’Neal DN, Best JD, Beilin LJ. Purified eicosapentaenoic and docosahexaenoic acids have differential effects on serum lipids and lipoproteins, LDL particle size, glucose, and insulin in mildly hyperlipidemic men. Am J Clin Nutr 2000;71:1085–94. 31. Siri PW, Krauss RM. Influence of dietary carbohydrate and fat on LDL and HDL particle distributions. Curr Atheroscler Rep 2005;7:455–9. 32. GISSI-Prevenzione Investigators. Dietary supplementation with n23 polyunsaturated fatty acids and vitamin E after myocardial infarction: results of the GISSI-Prevenzione trial. Lancet 1999;354:447–55. 33. Block RC, Harris WS, Reid KJ, Sands SA, Spertus JA. EPA and DHA in blood cell membranes from acute coronary syndrome patients and controls. Atherosclerosis 2008;197:821–8. 34. Yokoyama M, Origasa H, Matsuzaki M, Matsuzawa Y, Saito Y, Ishikawa Y, Oikawa S, Sasaki J, Hishida H, Itakura H, et al. Effects of eicosapentaenoic acid on major coronary events in hypercholesterolaemic patients (JELIS): a randomised open-label, blinded endpoint analysis. Lancet 2007;369:1090–8. 35. Lee KW, Blann AD, Lip GY. Effects of omega-3 polyunsaturated fatty acids on plasma indices of thrombogenesis and inflammation in patients post-myocardial infarction. Thromb Res 2006;118:305–12. 36. Kontush A, Chapman MJ. Antiatherogenic function of HDL particle subpopulations: focus on antioxidative activities. Curr Opin Lipidol 2010;21:312–8.
37. Tsai MY, Cao J. Lipoprotein subclasses: a changing landscape. Curr Cardiovasc Risk Rep 2008;2:23–8. 38. Johansson J, Carlson LA, Landou C, Hamsten A. High density lipoproteins and coronary atherosclerosis: a strong inverse relation with the largest particles is confined to normotriglyceridemic patients. Arterioscler Thromb 1991;11:174–82. 39. El Harchaoui K, Arsenault BJ, Franssen R, Despre´s JP, Hovingh GK, Stroes ES, Otvos JD, Wareham NJ, Kastelein JJ, Khaw KT, et al. High-density lipoprotein particle size and concentration and coronary risk. Ann Intern Med 2009;150:84–93. 40. Roche HM, Gibney MJ. The impact of postprandial lipemia in accelerating atherothrombosis. J Cardiovasc Risk 2000;7:317–24. 41. Gianturco SH, Bradley WA. Pathophysiology of triglyceride-rich lipoproteins in atherothrombosis: cellular aspects. Clin Cardiol 1999;22 (suppl):II7–14. 42. Chang CL, Seo T, Matsuzaki M, Worgall TS, Deckelbaum RJ. n23 fatty acids reduce arterial LDL-cholesterol delivery and arterial lipoprotein lipase levels and lipase distribution. Arterioscler Thromb Vasc Biol 2009;29:555–61. 43. Zhu J, Davidson M, Leinonen M, Saikku P, Gaydos CA, Canos DA, Gutman KA, Howard BV, Epstein SE; GOCADAN Study Investigators. Prevalence and persistence of antibodies to herpes viruses, Chlamydia pneumoniae and Helicobacter pylori in Alaskan Eskimos: The GOCADAN Study. Clin Microbiol Infect 2006;12:118–22.