Docosahexaenoic Acid Supplementation Improved Lipocentric but Not Glucocentric Markers of Insulin Sensitivity in Hypertriglyceridemic Men

Abstract

Background:

Increase in obesity and metabolic syndrome are associated with increases in insulin resistance (IR) and type 2 diabetes mellitus. Results from animal intervention studies and human epidemiological studies suggest that n-3 polyunsaturated fatty acids can prevent and reverse IR, but results from human intervention studies have varied. Results from some human and animal studies suggest that docosahexaenoic acid (22:6n-3; DHA) may be more effective than eicosapentaenoic acid (20:5n-3; EPA) in the prevention of IR.

Methods:

By using a placebo-controlled, parallel study design, we examined the effects of DHA supplementation (3 grams/day, 90 days) in the absence of EPA on glucocentric and lipocentric markers of IR in hypertriglyceridemic men (n=14–17/group).

Results:

DHA supplementation increased fasting plasma glucose concentration by 4.7% (P<0.05), but did not alter other indices of IR based on fasting (insulin and homeostasis model assessment of insulin resistance [HOMA-IR]) or postprandial insulin and glucose concentrations (areas under curves for insulin and glucose, Matsuda index). Glucose increased by 2.7% in the placebo group and was not significant; increases in glucose in the two groups did not differ from each other. DHA decreased circulating concentrations of several lipocentric markers of IR, including plasma concentrations of nonesterified fatty acids (13.0%), small, dense low-density lipoprotein (LDL) particles (21.7%), and ratio of tryglycerides to high-density lipoprotein cholesterol (TG/HDL-C) (34.0%) (P<0.05). None of the variables changed in the placebo group.

Conclusions:

Our results suggest that lipocentric markers of IR are more responsive to DHA supplementation than the glucocentric markers. Future studies with DHA in prediabetic subjects and direct measures of insulin sensitivity are needed.

Introduction

I nsulin resistance (IR) is a condition in which normal amounts of insulin fail to maintain normal blood glucose levels because of decreased responsiveness of muscle (glucose uptake), adipose tissue (glucose uptake and inhibition of lipolysis), and liver (inhibition of gluconeogenesis).^1,2 Thus, it is a reduced ability of insulin to exert its biological effects on its target tissues. IR is usually assessed by determining fasting or postprandial plasma concentrations of insulin and glucose (glucocentric markers).

A number of direct and surrogate methods are available to estimate glucocentric markers of IR.^3

–7 The two direct methods used to evaluate insulin-mediated glucose uptake (IMGU) are insulin suppression test (IST) and the euglycemic clamp. These direct methods are often time and labor intensive and require experienced operators. Therefore, other surrogate methods such as fasting plasma glucose (FPG), fasting plasma insulin (FPI), oral glucose tolerance test (OGTT), intravenous glucose-tolerance test (IVGTT), insulin tolerance test (ITT), homeostatic model assessment of insulin resistance (HOMA-IR), and the quantitative insulin sensitivity (IS) check index (QUICKI) are commonly used. The last two methods use fasting insulin and glucose concentrations to calculate IR or IS.

An equally important aspect of IR is the effects of insulin on lipid metabolism (lipocentric markers), which emphasizes that derangement in lipid metabolism can have devastating effects on glucose metabolism.⁸ Several physiological markers of lipid metabolism, such as the ratio between plasma triglycerides (TG) and high-density lipoprotein cholesterol (HDL-C), number of small, dense (sd) low-density lipoprotein (LDL) particles, and nonesterified fatty acids (NEFA) are associated with IR.^9

–15 Because increased inflammation is considered to be a major factor in the development of IR, some investigators believe that plasma concentrations of inflammatory markers are also useful in estimating the degree of IR.¹⁶

The amount and type of dietary fatty acids play a significant role in the development of IR. Saturated fatty acids (SFA), trans fatty acids (TFA), and n-6 polyunsaturated fatty acids (PUFA) increase inflammation and IR, whereas n-3 PUFA are believed to decrease inflammation and IR.^17
–19 Indications that n-3 PUFA may decrease IR came from epidemiological studies, inverse association between IR and tissue n-3 PUFA concentrations, and intervention studies with n-3 PUFA. Prior to the introduction of Western diets, which are characteristically high in n-6 PUFA, SFA, and TFA, the incidence of type 2 diabetes mellitus (T2DM) was much lower among Greenland Inuits and natives of Alaska than among the Danes and mainland residents of the United States.^20,21 Similarly, Japanese Islanders had a lower prevalence of noninsulin-dependent diabetes mellitus compared with their mainland counterparts.²² Results from two recent epidemiological studies conducted in China and Japan support a decrease in IR by increased intake of n-3 PUFA,^15,23 whereas those from three large epidemiological studies conducted within the mainland U.S. population found no association between n-3 PUFA intake and T2DM.¹⁹ A number of cross-sectional studies reported negative associations between IR and tissue concentrations of n-3 PUFA.^15,24

–27 Results from interventional studies have been variable; n-3 PUFA supplementation to nondiabetic human subjects improved IS in seven studies, but it did not change in ten studies.^17
–19 Inconsistencies among the results with fish oils could be because of a number of factors, including differences in the specific type of n-3 PUFA, their dose and duration of supplementation, overall fatty acid composition of the basal diet, health status of the study participants, and the methods used to evaluate IR.

Results from human studies with purified eicosapentaenoic acid (20:5n-3; EPA) and docosahexaenoic acid (22:6n-3; DHA) indicated that both of these fatty acids decreased plasma TG, but only DHA increased concentrations of plasma LDL-C and HDL-C, their mean particle sizes, and decreased concentrations of sd LDL particles.²⁸ Results from other studies suggest that DHA may be more effective than EPA in preventing IR.^29
–31 DHA supplementation improved IS in a study with obese children,³² but not in overweight and obese adults.³³ Further studies are needed to determine if DHA will improve IS in human subjects.

We chose to study the effects of DHA in the absence of EPA on risk factors for cardiovascular disease because it is the most potent fatty acid suppressor of hepatic lipogenesis³⁴ and is the major long-chain n-3 PUFA in human tissues. DHA can be readily retroconverted to EPA (which is not elongated to DHA in humans).³⁵ Previously we have reported that DHA supplementation decreased concentrations of inflammatory markers and improved lipid profiles in men with both normal as well as elevated TG.^36

–39 Here we report that DHA supplementation improved several lipocentric markers of IS, but did not alter the glucocentric markers that we tested.

Subjects and Methods

Subjects

Details regarding the study design and participants have been published previously.^36
–38 Moderately hyperlipidemic, but otherwise healthy, men (39–66 years old) participated in this study. Participants regularly taking antiinflammatory, antihypertensives, nonsulfonyl urea, or other medications were excluded. Also excluded were consumers of illegal substances, more than five drinks of alcohol per week, more than one fish meal per week, and those taking supplements of fish oil, krill oil, flaxseed oil, or vitamin C or E. Clinical chemistry and hematology panels for all qualified participants were in normal ranges with the exception of blood lipids. All selected participants had serum C-reactive protein (CRP) concentrations of 1–10 mg/L, fasting serum TGs concentrations of 150–400 mg/dL (1.70–4.53 mmol/L), total cholesterol <300 mg/dL (7.78 mmol/L), LDL-C <220 mg/dL (5.69 mmol/L), and body mass index (BMI) between 22 and 35 kg/m². Seventeen participants in each group completed the study.

Study design

The study protocol was approved by the Institutional Review Boards of the University of California Davis and the Veterans Affairs Medical Center, Mather, CA. It is listed in the government Clinical Trials (clinicaltrials.gov) by the identifier NCT00728338. The study design was a double-blind, placebo-controlled parallel study with two metabolic periods—baseline (first 8 days) and intervention (last 90 days). During the intervention period, participants supplemented their diets with either placebo or DHA capsules (Martek Biosciences Corporation, Columbia, MD). The DHA group received 7.5 grams/day of DHA oil, which is produced in the microalga Crypthecodinium cohinii. This oil contained the following fatty acids: 12:0 (3.0%), 14:0 (14.1%), 16:0 (14.2%), 18:1n-9 (23.8%), 22:6n-3 (40.9%), and other fatty acids (2.6%). Thus, the DHA group received approximately 3 grams/day of DHA, equivalent to 1.1% of daily energy intake. The placebo group received 7.5 grams/day of olive oil, which contained 16:0 (11.1%), 18:1n-9 (80.7%), 18:2n-6 (4.3%), and other fatty acids 3.9%. Participants continued to consume their regular diets and were instructed not to change their usual diets and activity levels throughout the study. Usual dietary intakes were estimated by three unannounced 24-h dietary recalls obtained by telephone using a multipass interview method during each of the metabolic periods; dietary intakes have been previously reported.³⁸

To provide uniformity in the composition of diets among subjects during blood draw days, the metabolic kitchen provided all three meals (pretest diet) the day before each blood draw. The energy intake was adjusted for body height, body weight, age, and estimated physical activity of the participants using the Mifflin–St. Jeor equation and appropriate activity factors. The pretest diet that was weighed and served on the day before each blood draw did not differ in composition or total energy intake between the two groups (not shown). For this day, mean energy intake for two groups was 10,450±240 kJ; mean intakes for fat, carbohydrates, and proteins were 82, 340, and 100 grams/day, respectively. Saturated, monounsaturated, and polyunsaturated fatty acids provided 11.1%, 10.1%, and 8.8% of the total energy.

Blood draws, lipids, and lipoproteins

Blood samples were drawn on study days −7 and 0 (baseline), day 45 (mid intervention), and days 84 and 91 (end of intervention), but most of the results reported here are based only on blood samples drawn before and after DHA supplementation (study days 0 and 91). After the subjects fasted for 12 h overnight, an indwelling cannula was inserted into the antecubital vein of the forearm and fasting blood samples were drawn. Postprandial blood samples were drawn at 2, 4, 6, and 8 h after the completion of a standardized breakfast. Plasma and serum were prepared by centrifugation at 4°C at 1,300×g for 10 min and stored either at −70°C or maintained unfrozen at 4°C (for lipoprotein sizing). Serum lipids and EDTA plasma lipoprotein sizes were determined by using nuclear magnetic resonance (NMR) spectroscopy, as previously reported.³⁸ The interassay coefficient of variance for determining the concentration of sd LDL particles and the mean size of LDL particles by NMR spectroscopy are 2.1% and 0.4%, respectively.⁴⁰ Serum insulin concentrations were determined by using an Immulite Chemiluminescent Analyzer (Siemens Diagnostics, Deerfield, IL); serum NEFA and plasma glucose (from Vacutainer tubes containing potassium oxalate and sodium fluoride) were determined using a clinical chemistry analyzer and clinical diagnostic kits (NEFA, Wako Diagnostics, Richmond, VA; glucose, Roche/Hitachi 902, Roche Diagnostics, Basel, Switzerland). Fasting serum glucose and insulin concentrations were used to calculate HOMA-IR, and postprandial (0, 2, 4, 6, 8 h) insulin and glucose concentrations were used to determine area under curve (AUC) for insulin and glucose and the Matsuda IS index.^6,41
–43

Statistical analysis

Prestudy power calculations were made using a one-sided t-test; an n=17 per group provided a power of 88.6% to detect significant treatment effects with group means of 1.0 and 0.9, standard deviation (SD) of 0.1, and alpha of 0.05. These calculations were based on primary outcome variables (TGs and concentrations of lipoprotein particles); glucocentric measures were not the primary outcome variables. The SAS proc mixed was used to fit a repeated measure, mixed model with a first-order autoregressive covariance structure among the repeated measures for data with more than two time points. Diet, time, and the interaction are the fixed effects and subjects within diets are the random effect as previously reported.³⁸ The areas under the curves were computed using the SAS ETS proc expand, which implements a trapezoidal rule approximation.⁴⁴ Results shown are the mean±standard error of the mean (SEM). P<0.05 is considered significant.

Results

Study subjects

Forty men entered the study, and 6 did not complete it. Three subjects dropped out from each of the two study groups (placebo, DHA). Two subjects dropped out because of time constraints (one in each group), two in the DHA group had a feeling of gas/bloating, and the other two (both in the placebo group) could not eat the test breakfast. Prestudy physical characteristics and fasting blood lipids for men who participated in the study are shown in Table 1. There were no significant differences between the two groups with respect to age, blood lipids, systolic or diastolic blood pressure, and heart rate; BMI was significantly greater in the placebo than in the DHA group (P<0.05).

Table 1.

Prestudy Physical and Fasting Biochemical Characteristics of Study Participants ^a

Variable	DHA group (n=17)	Placebo group (n=17)
Age, years	55.0±2.0	53.1±1.0
BMI, kg/m²	27.8±0.7^b	30.6±0.8^c
Triacylglycerols, mmol/L	2.55±0.16	2.71±0.19
Cholesterol, mmol/L	5.67±0.21	5.51±0.28
HDL-C, mmol/L	1.00±0.05	0.93±0.03
LDL-C, mmol/L	3.49±0.22	3.34±0.26
SBP, mmHg	128.6±2.93	128.8±3.6
DBP, mmHg	79.5±2.1	79.7±1.7
Heart rate, BPM	66.7±1.4	66.2±2.5

All values are mean±standard error of the mean (SEM). Numbers with different superscripts are significantly different between the two groups (P<0.05, nonpaired t-test).

DHA, docosahexaenoic acid; BMI, body mass index; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; SBP, systolic blood pressure; DBP, diastolic blood pressure; BPM, beats per minute.

Dietary intake and supplements

Intake of fiber, cholesterol, total energy, and percent of energy from carbohydrate, protein, fat, and the types of fat did not differ between the two study groups during the baseline period.³⁸ Furthermore, usual food intake did not change significantly between the baseline and intervention periods in both groups. Mean compliance with the consumption of supplements as determined by the number of unused capsules was >90% for the DHA group and >85% for the placebo group. Analysis of the plasma and red blood cell fatty acid compositions at the start and end of the study also confirmed compliance regarding the intake of dietary fats.³⁸

Results from this study regarding the effects of DHA on markers of inflammation and blood lipids have been previously published.^36
–38 Here we report the effects of DHA supplementation on glucocentric and lipocentric markers of IS.

DHA supplementation (3 grams/day, 90 days) increased plasma glucose concentration by 4.70±1.33% (mean±SEM, n=17) as compared to the corresponding values prior to its supplementation (P<0.05); corresponding change in the placebo group was 2.70±1.28% (mean±SEM, n=17), and the difference in the changes was not statistically significant (p day×treatments=0.14; Table 2). Fasting insulin concentrations and HOMA-IR based on fasting insulin and glucose concentrations did not change in both the placebo and DHA groups. Also unchanged were the AUC (0–8 h) for glucose and insulin as well as the product of the AUC for glucose and insulin concentrations. Matsuda index based on the 0 and 2 hours glucose and insulin values (Table 1) and 0–8 h values also did not change in either study group.

Table 2.

Effect of DHA on Markers of Insulin Sensitivity in Hypertriglyceridemic Men

	DHA		Placebo
Variable	Day 0	Day 91	Day 0	Day 91	P value day×treatment
Fasting glucose, mg/dL	101.7±2.5	106.7±3.5^*	99.7±2.1	102.3±2.4	0.14
Fasting insulin, μIU/mL	11.4±1.4	14.8±1.6	15.4±2.6	17.2±3.6	0.08
HOMA-IR	2.70±0.38	3.55±0.58	3.94±0.77	4.44±0.99	0.59
AUC (0–8 h) glucose, mg/dL	772±22	793±26	756±40	811±42	0.67
AUC (0–8 h) insulin, μIU/mL	292±42	361±57	422±93	522±113	0.56
AUC glucose×AUC insulin (0–8 h)	229,412±36,160	294,004±50,532	362,530±97,893	481,110±123,685	0.75
Matsuda index (0–2 h)	4.09±0.62	3.47±0.69	4.34±1.0	3.75±0.85	0.96
Matsuda index (0–8 h)	6.26±0.99	5.34±1.09	5.90±1.33	5.41±1.16	0.70
NEFA, μEq/L	460±30	400±30^*	420±30	420±30	0.18
sd LDL, nmol/L	1439±131	1127±113^*	1555±118	1417±110	0.19
Mean size LDL, nm	19.9±0.1	20.5±0.2^*	19.6±0.1	19.8±0.1	0.05
Apo B (μmol/L)	3.94±0.20	4.15±0.19	4.06±0.24	3.98±0.19	0.45
Apo CIII (μmol/L)	20.29±1.25	17.67±0.80^*	21.31±1.48	20.55±1.25	0.01

Hypertriglyceridemic men supplemented their diets with DHA (3 grams/day) or placebo for 90 days. Fasting and postprandial (0–8 h) circulating markers of insulin resistance (IR) were determined before and after the supplements. Data shown are mean±standard error of the mean (SEM) (n=14–17/group). Within group comparisons [analysis of variance (ANOVA)] of day 0 and day 91 showed significant (P<0.05) change for variables bearing an asterisk (^*). AUC=area under curve for 0, 2, 4, 6, and 8 h following a test meal. Reference values in healthy subjects for NEFA, sd LDL, and LDL size are 239–843 μEq/L⁵⁴, 877–1793 nmol/L, and 19.8–21.9 nm,¹⁰ respectively. Data for LDL and Apo proteins are taken from reference 36, which also includes the data regarding other lipids and lipoproteins.

DHA, docosahexaenoic acid; HOMA-IR, homeostasis model assessment of insulin resistance; AUC, area under the curve; NEFA, nonesterified fatty acids; sd LDL, small, dense low-density lipoprotein; Apo B, apolipoprotein B; Apo CIII, apolipoprotein CIII.

The lipocentric markers of IS (serum NEFA and plasma sd LDL; Table 2) and the ratio between TG and HDL-C (Fig. 1) were all significantly decreased by DHA supplementation (P<0.05), but did not change in the placebo group. DHA supplementation for 90 days decreased the concentrations of NEFA, sd LDL, and TG/HDL-C ratio by 13.0%, 21.7%, and 34.0%, respectively. It increased the mean LDL particle size by 3% (0.6 nm, P<0.05).

FIG. 1.

Effect of docosahexaenoic acid (DHA) supplementation (3 grams/day, 90 days) on plasma triglyceride/high-density lipoprotein cholesterol (TG/HDL-C) ratio (top left), placebo supplement (top right), and the area under the curve (AUC, bottom center) to hypertriglyceridemic men. Data are mean±standard error of the mean (SEM) (n=17/group). Bars with different letters for DHA group are significantly different (P<0.05). P=0.0005 for interaction between day and treatment for AUC. Reference ranges for triglycerides, HDL-C, and triglyceride: HDL-C ratios are 37–877 mg/dL, 17–87 mg/dL, and 0.8–28.3,¹⁰ respectively.

Discussion

Results from our study with hypertriglyceridemic men show that DHA supplementation improved several markers of IS based on circulating lipids; it did not change markers based on the fasting or postprandial glucose and insulin concentrations, except a for 4.7% increase in fasting glucose (P<0.05). Increase in fasting glucose (2.7%) in the placebo group was not significant, and it did not differ from the increase in the DHA group (4.7%). Furthermore, HOMA-IR did not change in both groups. Together, these findings suggest that the small increase in plasma glucose by DHA may not be clinically relevant. Our results dealing with the lipocentric markers of IS are consistent with those reported with fish oils^9

–14 or DHA.^{33,38,45

–49} Our findings regarding the effects of DHA on glucocentric markers of IR are consistent with those from a DHA study with overweight and obese adults where DHA failed to improve IS,³³ but differ from those of study with prediabetic children where DHA improved IS.³² Both of these studies collected several blood samples within 2 h after an oral glucose load, but the dose and duration of DHA used and subject characteristics were quite different. Our study design differed from the designs of these two studies in several ways, but two major differences were that we used a meal test instead of OGTT and we collected the first blood samples 2 h after the intake of a meal and they collected samples every 30 min.

The lack of change in markers based on glucose and insulin concentrations in our study may be real, but most likely was due to the study design and the methods used. Our inability to detect enhanced IS following DHA supplementation may be because of several reasons. Fasting glucose and insulin are not the best markers to detect IR because they represent only hepatic IR and do not provide precise estimates of IMGU. Fasting glucose does not reveal IR until the pancreatic β-cells cannot keep up with the increased demand for insulin to maintain normal blood glucose levels. Surrogate estimates of IMGU based on FPI and FPG account for no more than 13% of the variability in insulin action in normal-weight groups, 30% in overweight groups, and 37% in obese groups.⁴ In addition to the fasting samples, we also collected postprandial blood samples at 2, 4, 6, and 8 h after a test meal. Serum insulin and glucose concentrations did not differ between the two groups at all these postprandial time points (not shown); the AUC for both insulin and glucose concentrations also did not change (Table 1). Because the focus of this study was on blood lipids and not on glucose or insulin, we did not collect blood samples earlier than 2 h after the meal. It is possible we would have found differences in insulin concentrations if we had collected postprandial blood samples earlier than 2 h or if we had used the euglycemic clamp method.

Participants in our study had normal levels of fasting glucose and insulin. In a previous study, n-3 long-chain polyunsaturated fatty acids (LCPUFA; 70% DHA and 30 EPA%) supplementation improved IS in obese patients with increased IR, but not in the group with normal IS.⁵⁰ In another weight loss study with overweight women, n-3 PUFA supplementation improved IS in subjects with an inflammatory phenotype, but did not change in those without elevated markers of inflammation.⁵¹ DHA supplementation decreased concentrations of several markers of inflammation in our study, but they were not accompanied by changes in glucocentric markers of IR.³⁶ Our results may differ from those of the previously published study because of differences in sex, weight loss, and the degree of inflammation.

One of the physiologic lipocentric markers of improved IS is a decrease in the ratio between the plasma TG and HDL-C concentrations.^11,13 Within 45 days of fatty acid supplementation, AUC values (0–8 h) for the ratio between TG and HDL-C decreased 34% and 7% in the DHA and placebo groups, respectively (Fig. 1), and this was maintained for the next 45 days. These differences resulted from a decrease in TG and an increase in HDL-C in the DHA group. Other lipocentric markers suggesting improved IS by DHA in our study include a decrease in sd LDL particles and NEFA and an increase in large LDL particles.³⁸

We, as well as others, have demonstrated that DHA reversed IR and nonalcoholic fatty liver disease (NAFLD) in animal models.^31,52 Furthermore, 4-OH DHA was a more potent stimulator of peroxisome proliferator-activated receptor γ (PPARγ) than pioglitazone, and it was nearly as effective as pioglitazone in preventing diabetes in rodent models.⁵³ We believe our inability to detect improvements in glucocentric markers of IS was most likely due to our study design, insensitive methods, and the health status of the study participants.

Our study had several limitations. The number of subjects who participated in the study was based on lipocentric markers; the glucocentric markers were not the primary end points. Plasma concentrations of glucocentric markers for IR were within the normal ranges for healthy subjects, and it may not be possible to improve them further with DHA. The methods used to evaluate glucocentric markers of IR lacked sensitivity. We may have found an effect of DHA on IR if we had used subjects with increased IR and used the euglycemic clamp, IST, or repeated blood draws after a glucose load test.

Regardless of the above limitations, our results demonstrate that the lipocentric surrogate markers are more sensitive than the glucocentric surrogate markers of IR. They emphasize the need to assess glucocentric markers directly (euglycemic clamp, IST) and also to consider the metabolic relevance, reproducibility, and sensitivity of the methods used when comparing results between different studies. Future studies with prediabetic subjects and improved methods of IS are needed to determine if DHA will improve IS.

Footnotes

Acknowledgments

We thank the study participants for their time and commitment, Dr. Ellen Bonnel and her staff for coordination of the study, and the BioAnalytical Support Laboratory for clinical chemistry analyses.

Author Disclosure Statement

Funds for this study were provided by USDA and DHA was donated by Martek Biosciences Corporation, a division of DSM Nutritional Products, Columbia, MD 21045. No competing financial interests exist.

Darshan S. Kelley, Bruce E. Mackey, David Siegel, and Leslie R. Woodhouse were responsible for conception and planning of study, data interpretation, and preparation of manuscript; Yuriko Adkins performed sample analysis and preparation of the manuscript; Arthur Swislocki was responsible for interpretation of data and preparation of the manuscript.

References

Marchesini

, Brizi

, Morselli-Labate

et al. Association of nonalcoholic fatty liver disease with insulin resistance. Am J Med, 1999; 107:450–455.

Moller

, Flier

. Insulin resistance—mechanisms, syndromes, and implications. N Engl J Med, 1991; 325:938–948.

Borai

, Livingstone

, Ferns

. The biochemical assessment of insulin resistance. Ann Clin Biochem, 2007; 44:324–342.

Kim

, Abbasi

, Reaven

. Impact of degree of obesity on surrogate estimates of insulin resistance. Diabetes Care, 2004; 27:1998–2002.

Lee

, Muniyappa

, Yan

et al. Comparison between surrogate indexes of insulin sensitivity/resistance and hyperinsulinemic euglycemic glucose clamps in rhesus monkeys. Endocrinology, 2011; 152:414–423.

Matsuda

. Measuring and estimating insulin resistance in clinical and research settings. Nutr Metab Cardiovasc Dis, 2010; 20:79–86.

Piche

, Arcand-Bosse

, Despres

et al. What is a normal glucose value? Differences in indexes of plasma glucose homeostasis in subjects with normal fasting glucose. Diabetes Care, 2004; 27:2470–2477.

McGarry

. Banting lecture 2001: Dysregulation of fatty acid metabolism in the etiology of type 2 diabetes. Diabetes, 2002; 51:7–18.

Burns

, Lee

, Arslanian

. In vivo insulin sensitivity and lipoprotein particle size and concentration in black and white children. Diabetes Care, 2009; 32:2087–2093.

10.

Fan

, Liu

, Hoffman

et al. Triglyceride/high-density lipoprotein cholesterol ratio: a surrogate to predict insulin resistance and low-density lipoprotein cholesterol particle size in nondiabetic patients with schizophrenia. J Clin Psychiatry, 2011; 72:806–812.

11.

Laws

, King

, Haskell

et al. Relation of fasting plasma insulin concentration to high density lipoprotein cholesterol and triglyceride concentrations in men. Arterioscler Thromb, 1991; 11:1636–1642.

12.

Laws

, Reaven

. Evidence for an independent relationship between insulin resistance and fasting plasma HDL-cholesterol, triglyceride and insulin concentrations. J Intern Med, 1992; 231:25–30.

13.

McLaughlin

, Abbasi

, Cheal

et al. Use of metabolic markers to identify overweight individuals who are insulin resistant. Ann Intern Med, 2003; 139:802–809.

14.

McLaughlin

, Deng

, Gonzales

et al. Insulin resistance is associated with a modest increase in inflammation in subcutaneous adipose tissue of moderately obese women. Diabetologia, 2008; 51:2303–2308.

15.

Huang

, Wahlqvist

, Xu

et al. Increased plasma n-3 polyunsaturated fatty acid is associated with improved insulin sensitivity in type 2 diabetes in China. Mol Nutr Food Res, 2010; 54,Suppl 1:S112–S119.

16.

Savage

, Petersen

, Shulman

. Mechanisms of insulin resistance in humans and possible links with inflammation. Hypertension, 2005; 45:828–833.

17.

Djousse

, Gaziano

, Buring

et al. Dietary omega-3 fatty acids and fish consumption and risk of type 2 diabetes. Am J Clin Nutr, 2011; 93:143–150.

18.

Fedor

, Kelley

. Prevention of insulin resistance by n-3 polyunsaturated fatty acids. Curr Opin Clin Nutr Metab Care, 2009; 12:138–146.

19.

Kaushik

, Mozaffarian

, Spiegelman

et al. Long-chain omega-3 fatty acids, fish intake, and the risk of type 2 diabetes mellitus. Am J Clin Nutr, 2009; 90:613–620.

20.

Mouratoff

, Carroll

, Scott

. Diabetes mellitus in Athabaskan Indians in Alaska. Diabetes, 1969; 18:29–32.

21.

Kromann

, Green

. Epidemiological studies in the Upernavik district, Greenland. Incidence of some chronic diseases 1950–1974. Acta Med Scand, 1980; 208:401–406.

22.

Kagawa

, Nishizawa

, Suzuki

et al. Eicosapolyenoic acids of serum lipids of Japanese islanders with low incidence of cardiovascular diseases. J Nutr Sci Vitaminol (Tokyo), 1982; 28:441–453.

23.

Muramatsu

, Yatsuya

, Toyoshima

et al. Higher dietary intake of alpha-linolenic acid is associated with lower insulin resistance in middle-aged Japanese. Prev Med, 2010; 50:272–276.

24.

Nigam

, Frasure-Smith

, Lesperance

et al. Relationship between n-3 and n-6 plasma fatty acid levels and insulin resistance in coronary patients with and without metabolic syndrome. Nutr Metab Cardiovasc Dis, 2009; 19:264–270.

25.

Thorseng

, Witte

, Vistisen

et al. The association between n-3 fatty acids in erythrocyte membranes and insulin resistance: The Inuit Health in Transition Study. Int J Circumpolar Health, 2009; 68:327–336.

26.

Burrows

, Collins

, Garg

. Omega-3 index, obesity and insulin resistance in children. Int J Pediatr Obes, 2011; 6:e532–e539.

27.

Haugaard

, Madsbad

, Hoy

et al. Dietary intervention increases n-3 long-chain polyunsaturated fatty acids in skeletal muscle membrane phospholipids of obese subjects. Implications for insulin sensitivity. Clin Endocrinol (Oxf), 2006; 64:169–178.

28.

Cottin

, Sanders

, Hall

. The differential effects of EPA and DHA on cardiovascular risk factors. Proc Nutr Soc, 2011; 70:215–231.

29.

Egert

, Fobker

, Andersen

et al. Effects of dietary alpha-linolenic acid, eicosapentaenoic acid or docosahexaenoic acid on parameters of glucose metabolism in healthy volunteers. Ann Nutr Metab, 2008; 53:182–187.

30.

Gotoh

, Nagao

, Onoda

et al. Effects of three different highly purified n-3 series highly unsaturated fatty acids on lipid metabolism in C57BL/KsJ-db/db mice. J Agric Food Chem, 2009; 57:11047–11054.

31.

Vemuri

, Kelley

, Mackey

et al. Docosahexaenoic acid (DHA) but not eicosapentaenoic acid (EPA) prevents trans-10, cis-12 conjugated linoleic acid (CLA)-induced insulin resistance in mice. Metabol Syndr Related Dis, 2007; 5:315–322.

32.

Nobili

, Bedogni

, Alisi

et al. Docosahexaenoic acid supplementation decreases liver fat content in children with non-alcoholic fatty liver disease: double-blind randomised controlled clinical trial. Arch Dis Child, 2011; 96:350–353.

33.

Neff

, Culiner

, Cunningham-Rundles

et al. Algal docosahexaenoic acid affects plasma lipoprotein particle size distribution in overweight and obese adults. J Nutr, 2011; 141:207–213.

34.

Jump

. N-3 polyunsaturated fatty acid regulation of hepatic gene transcription. Curr Opin Lipidol, 2008; 19:242–247.

35.

Brenna

, Salem

Jr. , Sinclair

et al. Alpha-linolenic acid supplementation and conversion to n-3 long-chain polyunsaturated fatty acids in humans. Prostaglandins Leukot Essent Fatty Acids, 2009; 80:85–91.

36.

Kelley

, Siegel

, Fedor

et al. DHA supplementation decreases Serum C-reactive protein and other markers of inflammation in hypertriglyceridemic men. J Nutr, 2009; 139:495–501.

37.

Kelley

, Siegel

, Vemuri

et al. Docosahexaenoic acid supplementation decreases remnant-like particle-cholesterol and increases the (n-3) index in hypertriglyceridemic men. J Nutr, 2008; 138:30–35.

38.

Kelley

, Siegel

, Vemuri

et al. Docosahexaenoic acid supplementation improves fasting and postprandial lipid profiles in hypertriglyceridemic men. Am J Clin Nutr, 2007; 86:324–333.

39.

Kelley

, Taylor

, Nelson

et al. Dietary docosahexaenoic acid and immunocompetence in young healthy men. Lipids, 1998; 33:559–566.

40.

Jeyarajah

, Cromwell

, Otvos

. Lipoprotein particle analysis by nuclear magnetic resonance spectroscopy. Clin Lab Med, 2006; 26:847–870.

41.

Choukem

, Gautier

. How to measure hepatic insulin resistance? Diabetes Metab, 2008; 34:664–673.

42.

Lee

, Muniyappa

, Yan

et al. Comparison between surrogate indexes of insulin sensitivity and resistance and hyperinsulinemic euglycemic clamp estimates in mice. Am J Physiol Endocrinol Metab, 2008; 294:E261–E270.

43.

Matsuda

, DeFronzo

. Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care, 1999; 22:1462–1470.

44.

SAS. SAS OnlineDoc 9.2. Cary, NC: SAS Institute Inc., 2009.

45.

Mori

, Burke

, Puddey

et al. 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; 71:1085–1094.

46.

Woodman

, Mori

, Burke

et al. Effects of purified eicosapentaenoic and docosahexaenoic acids on glycemic control, blood pressure, and serum lipids in type 2 diabetic patients with treated hypertension. Am J Clin Nutr, 2002; 76:1007–1015.

47.

Engler

, Engler

, Malloy

et al.

Effect of docosahexaenoic acid on lipoprotein subclasses in hyperlipidemic children (the EARLY study)

Am J Cardiol, 2005; 95:869–871.

48.

Rambjor

, Walen

, Windsor

et al. Eicosapentaenoic acid is primarily responsible for hypotriglyceridemic effect of fish oil in humans. Lipids, 1996; 31,Suppl:S45–S49.

49.

Theobald

, Chowienczyk

, Whittall

et al. LDL cholesterol-raising effect of low-dose docosahexaenoic acid in middle-aged men and women. Am J Clin Nutr, 2004; 79:558–563.

50.

Browning

, Krebs

, Moore

et al. The impact of long chain n-3 polyunsaturated fatty acid supplementation on inflammation, insulin sensitivity and CVD risk in a group of overweight women with an inflammatory phenotype. Diabetes Obes Metab, 2007; 9:70–80.

51.

Krebs

, Browning

, McLean

et al. Additive benefits of long-chain n-3 polyunsaturated fatty acids and weight-loss in the management of cardiovascular disease risk in overweight hyperinsulinaemic women. Int J Obes (Lond), 2006; 30:1535–1544.

52.

Rossmeisl

, Jelenik

, Jilkova

et al. Prevention and reversal of obesity and glucose intolerance in mice by DHA derivatives. Obesity (Silver Spring), 2009; 17:1023–1031.

53.

Yamamoto

, Itoh

, Abe

et al. Identification of putative metabolites of docosahexaenoic acid as potent PPARgamma agonists and antidiabetic agents. Bioorg Med Chem Lett, 2005; 15:517–522.

54.

Wallach

. Normal blood levels (reference values) Wallach

. Interpretation of Diagnostic Tests—A Synopsis of Laboratory Medicine, 5th. Boston: Little, Brown and Company, 1992; 1–29.