AJCN North Carolina Research Campus
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Tables
Right arrow An erratum has been published
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mostad, I. L
Right arrow Articles by Grill, V.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mostad, I. L
Right arrow Articles by Grill, V.
Agricola
Right arrow Articles by Mostad, I. L
Right arrow Articles by Grill, V.
American Journal of Clinical Nutrition, Vol. 84, No. 3, 540-550, September 2006
© 2006 American Society for Nutrition

ORIGINAL RESEARCH COMMUNICATION

Effects of n–3 fatty acids in subjects with type 2 diabetes: reduction of insulin sensitivity and time-dependent alteration from carbohydrate to fat oxidation 1,3

Ingrid L Mostad, Kristian S Bjerve, Marit R Bjorgaas, Stian Lydersen and Valdemar Grill

1 From the Department of Cancer Research and Molecular Medicine, Faculty of Medicine (ILM, MRB, and VG), and the Unit for Applied Clinical Research (SL), Norwegian University of Science and Technology, Trondheim, Norway, and the Department of Medical Biochemistry (KSB), the Department of Medicine, Division of Endocrinology (MRB and VG), and the Department of Clinical Service, Division of Clinical Nutrition (ILM), St Olavs Hospital, Trondheim, Norway

2 This project was financed with the aid of EXTRA funds from the Norwegian Foundation for Health and Rehabilitation. The study was also supported by Peter Möller AS, Novo Nordisk, Abbot Norge AS, and the Norwegian Diabetes Association.

3 Reprints not available. Address correspondence to IL Mostad, Department of Cancer Research and Molecular Medicine, Faculty of Medicine, Norwegian University of Science and Technology, Olav Kyrres gt 3, MTFS, N-7489 Trondheim, Norway. E-mail: ingrid.l.mostad{at}ntnu.no.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Background: Effects of fish oil supplements on metabolic variables are insufficiently clarified in type 2 diabetes.

Objective: We aimed to investigate short-term (1 wk) and longer-term (9 wk) effects of n–3 fatty acids.

Design: Twenty-six subjects with type 2 diabetes without hypertriacylglycerolemia participated in a double-blind controlled study. Median intake in the intervention group was 17.6 mL fish oil/d (1.8 g 20:5n–3, 3.0 g 22:6n–3, and 5.9 g total n–3 fatty acids). The control group received 17.8 mL corn oil/d (8.5 g 18:2n–6).

Results: Plasma phospholipid 20:5n–3 and 22:6n–3 increased, whereas 18:2n–6 decreased, in the fish oil group compared with the corn oil group after 1 wk. The two n–3 fatty acids also increased in adipose tissue biopsy samples taken after 9 wk in the fish oil group. Glucose concentrations (home-monitored) were {approx}1 mmol/L higher in the fish oil group than in the corn oil group at the end of the intervention (P = 0.035). Glucose utilization measured by using an isoglycemic clamp was lowered in the fish oil group compared with that in the corn oil group at the end of the intervention (P = 0.049), whereas glucagon-stimulated C-peptide tended to increase (P = 0.078). The fish oil group utilized less fat for oxidation after 1 wk, with a change to more fat and less carbohydrate oxidation after 9 wk (P = 0.040), than did the corn oil group.

Conclusion: A high intake of fish oil moderately increases blood glucose and decreases insulin sensitivity in persons with type 2 diabetes without hypertriacylglycerolemia and alters carbohydrate and fat utilization in a time-dependent manner.

Key Words: Insulin sensitivity • insulin secretion • type 2 diabetes • fish oil • n–3 fatty acids • indirect calorimetry • energy metabolism • fat oxidation • carbohydrate oxidation


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Research over past decades has shown that marine n–3 fatty acids have therapeutic potential in preventive medicine, including in cardiovascular disease and cancer (1, 2). In Norway, fish oil is recommended to ensure vitamin D supply (3) and to prevent thrombosis (4) and cardiovascular disease (5). The evidence in subjects with type 2 diabetes, however, is equivocal as to the benefits of n–3 fatty acid supplementation in the diet. In contrast with animal studies (6), none of the controlled intervention studies in subjects with type 2 diabetes found better insulin sensitivity after supplementation with n–3 fatty acids (6-8). Deterioration of metabolic control was found in some (9-14) but not all (6, 8, 15-22) studies. Impaired metabolic control has been explained by increased fat intake per se (12), unfortunate choice of placebo (olive oil) (10, 13), or decreased insulin secretion (10-12, 14). Most previous studies were small (9 of 15 studies included 8–16 subjects), and the larger studies did not include detailed investigations of insulin sensitivity or insulin secretion. Thus, the available data appear insufficient to formulate clear-cut recommendations for supplementation with n–3 fatty acids in subjects with type 2 diabetes.

Several confounding factors may have contributed to the previous discrepant results. First, most studies used a crossover design, which has a risk of carry-over effects. Second, the presence or absence of hypertriacylglycerolemia could be important, because n–3 fatty acids reduce hypertriacylglycerolemia, which might affect insulin sensitivity. Third, complete blinding of study medication was not documented in most previous studies. Fourth, unequal duration of intervention might also have contributed to the discrepant results. n–3 Fatty acids may have different biological effects in the short and long term, because incorporation of n–3 fatty acids into adipocytes is slow (23), whereas incorporation into plasma fractions is rapid (24). To our knowledge, the time-course of n–3 fatty acid effects has not been systematically investigated in humans.

We performed a randomized, double-blind controlled study in subjects with type 2 diabetes without hypertriacylglycerolemia to clarify some of the mentioned uncertainties. Effects were assessed after 1 and 9 wk of intervention to illuminate the time course of n–3 fatty acids. We assessed metabolic control, insulin sensitivity, insulin secretion, and energy metabolism by use of home glucose monitoring, isoglycemic clamp, C-peptide glucagon test, and indirect calorimetry. To facilitate detection of time-course effects, we chose to intervene with a high dose (6 g) of n–3 fatty acids.


   SUBJECTS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Study population and design
Inclusion and exclusion criteria
Subjects between 40 and 75 y of age were recruited through advertisements in the local newspaper. Type 2 diabetes was defined by clinical criteria and by the absence of antibodies to glutamic acid decarboxylase. Participants had never used supplement with fish oil or marine n–3 fatty acids or discontinued such supplements for ≥6 mo before baseline. Exclusion criteria were insulin treatment, hypertriacylglycerolemia (>2.2 mmol/L), proliferative retinopathy, pregnancy or lactation, allergy to fish or citrus, smoking, alcoholism, and congestive heart failure or other serious diseases affecting the ability of the subject to participate.

General design
The study had a parallel design. Subjects were screened on the basis of their medical history and a physical examination. The period from inclusion to baseline (wk 0) varied between 1 and 12 mo. Randomization was performed by the method of minimization (25), with the factors body mass index, glycated hemoglobin, the use of oral hypoglycemic medication, sex, age, fasting C-peptide concentration, and fatty fish intake, which were weighted as 4, 3, 3, 2, 2, 2, and 1, respectively (26). The intervention lasted for 9 wk and effects were registered after 1 and 9 wk.

Study population
Of the 27 subjects included (13 M, 14 F), 14 (6 M, 8 F) were randomly assigned to receive corn oil and 13 (7 M, 6 F) to receive fish oil. One subject assigned to the fish oil group left the study after 3 d of intervention because of noncompliance with the protocol. Data from this subject are excluded from all results.

Details on intervention
Subjects were instructed to adhere to their usual diet, intake of alcohol (if any), and usual level of physical activity throughout the intervention period. The intervention consisted of a daily intake of 20 mL fish oil enriched with n–3 fatty acids, given as liquid. We used liquid instead of capsules because liquid is the most comfortable way of ingesting a high dose. Furthermore, in Norway, fish oil in liquid form is traditionally ingested during the winter. Control subjects ingested equal amounts of corn oil. The oils were distributed in identical bottles by Peter Möller, Division of Orkla ASA (Oslo, Norway). The composition of the 2 types of oil is given in Table 1Go. The fish oil contained no vitamin D, but vitamin A had not been completely removed. Therefore, vitamin A was added to the corn oil preparation to match the concentration remaining in the fish oil. Both oil preparations were flavored with lemon and were supplemented with a mixture of natural tocopherols and vitamin E, the latter in the form of all-rac-{alpha}-tocopherol (Table 1Go). To ensure endogenous regeneration of vitamin E, subjects were instructed to ingest 60 mg ascorbic acid/d (vitamin C) together with the oil.


View this table:
[in this window]
[in a new window]
  		TABLE 1 Mean composition of the corn oil and fish oil supplements 1

 
The subjects received disposable premarked 20-mL dose cups and administered the daily dosage of oils and vitamins at home. The oil bottles were weighed at baseline and at the end of the study. The weight difference was used to calculate the amount ingested during the study period, using {rho} = 0.925 as the oil density. Remaining vitamin C tablets were collected and counted.

During the intervention, the subjects were discouraged from discussing the category of oil with the investigators. Subjects were tested on different days; hence, the opportunities for participants to share experiences were limited. After completing the study, the subjects were asked which of the 2 preparations they thought they had ingested. They sent their answers to the randomization office in a preaddressed envelope. Nine subjects had no opinion on which preparation they had ingested. Thirteen subjects had an opinion based on taste, but of these, 6 were wrong. Four subjects had an opinion independent of taste, and 3 of these turned out to be correct. In all, 7 of 14 subjects in the corn oil group suspected that they had ingested corn oil, whereas in the fish oil group, only 3 of 12 subjects suspected that they had ingested fish oil. From these data, we conclude that the study was successfully blinded.

Ethics
The local ethics committee approved the study protocol. All participants gave written informed consent. No reward was given to the study subjects, except for individual counseling about diet and lifestyle after completion of the study.

Tests
Test days
Measurements in the hospital were performed at baseline and after 1 and 9 wk of intervention. At each of these 3 stages, we used 2 consecutive days for testing. For each individual subject, we used the same weekdays for testing at baseline and after 1 and 9 wk of intervention. Before each test day, the subjects fasted 12 h (and abstained from alcohol for 24 h). They were told not to engage in strenuous exercise on the day before testing. In the morning on test days, the subjects traveled for <1 h to arrive at the hospital at 0800. On the first day of testing, blood and urine samples and anthropometric data were collected. Insulin secretion was measured by the C-peptide glucagon test. On the second day of testing, isoglycemic hyperinsulinemic clamps were performed mostly as described (27), with indirect calorimetry measurements (28, 29) included. At baseline and after 9 wk of intervention, dual-energy X-ray absorptiometry (DXA) scanning was performed. Furthermore, adipose tissue biopsy was performed and diet was recorded.

C-peptide glucagon tests
Serum concentrations of C-peptide and insulin were measured 5, 6, and 7 min after intravenous injection of 1 mg glucagon. The peak C-peptide response (ie, the increment above that recorded before injection) was used as the main variable of insulin secretion.

Isoglycemic hyperinsulinemic clamps
After indirect calorimetry measurements were made in the basal resting state (see below) and before the start of the clamp, a cannula (Venflon; Viggo, Helsingborg, Sweden) was inserted into a vein dorsally on one hand for blood sampling. During the clamp procedure, the arm used for blood sampling was warmed by different remedies (such as plastic gloves filled with warm water) to "arterialize" the blood. A second cannula for infusions was inserted contralaterally into a hand vein or antecubital vein. Infusion of insulin and glucose was then started. Insulin [1.5 mL insulin Actrapid (Novo Nordisk A/S, Bagsværd, Denmark); 100 IU/mL in 500 mL isotonic saline to which 2 mL of the subject's blood was added, ie, 0.3 IU/mL] was infused at a rate of 40 mU · min–1 · m–2 for 2 h. A 20% solution of glucose was infused at a variable rate. Measurements of blood glucose taken every 5 min served to adjust the infusion rate of glucose. The glucose infusion aimed to maintain the concentration of fasting glucose (measured at baseline). Insulin sensitivity was calculated as the amount of glucose [mg · min–1 · kg–1 lean body mass (LBM)] needed to maintain the baseline concentration of fasting glucose during the last 40 min of each clamp procedure. Amounts of glucose in urine samples were negligible.

Indirect calorimetry
Indirect calorimetry (Deltatrac II metabolic monitor; Datex-Ohmeda Division, Helsinki, Finland) was performed immediately before and during part of each isoglycemic clamp. Oxygen consumption (mL O2/min) and carbon dioxide production (mL CO2/min) were measured while the subjects breathed through a ventilated canopy hose that was coupled to the Deltatrac gas exchange monitor. Measurements of respiratory gas exchange were performed for 30 min in the basal resting state and were then repeated for 30 min of the last hour of the clamp. Calculations of results included the energy production rate (EPR, kcal/24 h) and the respiratory quotient (RQ), ie, the ratio between CO2 produced in mL/min (VCO2), and O2 consumed in mL/min (VO2). Calculations of protein oxidation were based on each subject's urinary loss of nitrogen (UN; mg/24 h). The latter was calculated from the output of carbamide and creatinine in urine produced during the preceding overnight period (ie, 9 h) as well as in urine collected during each clamp period. The following equation was used (29):

Formula 1(1)
and UN was extrapolated from 9 to 24 h.

Other variables and assays
Anthropometric measurements
Height (in cm) was measured to one decimal. Weight (in kg) was measured to one decimal with the subjects wearing indoor clothing and without shoes. The same electronic scale was used on all occasions. Waist circumference was measured with the subject in underwear, with shoulders relaxed and arms hanging alongside. Blood pressure was measured after the relaxing hour of indirect calorimetry and before the clamp. Body composition was assessed by DXA scanning (Hologic QDR 4500 A; Hologic Inc, Bedford, MA). Scans were reviewed and analyzed by using Hologic software 9.02.b (1996). Total body weight given by the DXA scanning was used to calculate EPR. The LBM recorded at baseline was used both for calculations at baseline and after 1 wk of intervention (because scanning was performed only at baseline and after 9 wk).

Diet
Intake of food and beverages was recorded with the use of frequency questionnaires at baseline and at 9 wk. The answers were analyzed with regard to intake of energy and nutrients. The results were computed by using a food database (AKF96) and software systems (BEREGN) developed at the Department of Nutrition Research, University of Oslo. The food database was mainly based on the official Norwegian food table (30). All fatty acids in the diet were expressed in g/d after estimating intake on the basis of choices of portion sizes and frequencies of food intake in the questionnaires, and the respective percent of energy was calculated for fatty acid groups: saturated fatty acids (SFAs), monounsaturated fatty acids (MUFAs), and n–6 and n–3 fatty acids.

Physical activity
The level and intensity of the subjects' usual physical activity was recorded by questionnaire at screening.

Home glucose monitoring
The subjects performed home glucose measurements on day 1 in week 0 and then on the same weekday of weeks 1, 3, 5, 6, and 8. On the measurement days, the subjects monitored their blood glucose 5 times: fasting before breakfast, before lunch, before dinner, 2 h after dinner, and at bedtime. All subjects used the same glucose measuring device (Precision; Abbott Laboratories, MediSense Products, Bedford, MA).

Plasma total phospholipid fatty acids
These were analyzed as previously described (31). The individual fatty acids were measured as mg/L.

Adipose tissue biopsies
Subcutaneous adipose tissue was obtained from one buttock by needle aspiration (Microlance 3, 1.6 x 40 mm; Becton Dickinson and Company Ltd, Drogheda, Ireland) as described previously (32). The specimen was placed on ice and frozen immediately at –80 °C, pending later analyses. The composition of fatty acids was determined by a modified version of a method previously described (33, 34).

Assays
Before and during the clamp procedures, blood and urinary glucose were measured with a glucose analyzer (YSI; Yellow Springs Instrument, Yellow Springs, OH). Serum samples were centrifuged (2000 x g, 12 min, 20 °C) and were analyzed immediately or kept at –80 °C for later analyses. Serum glucose, glycated hemoglobin, triacylglycerols, total cholesterol, HDL cholesterol, urine albumin, carbamide, and creatinine were measured by use of standard laboratory techniques. LDL cholesterol was calculated by use of the Friedewald formula (35). Free fatty acids (FFAs) were analyzed in EDTA-containing plasma or in serum. Samples were centrifuged immediately (1500 x g, 15 min, 20 °C) and frozen and kept at –80 °C pending later analyses by an enzymatic colorimetric method (NEFA-C kit; Wako Pure Chemical Industries Ltd, Osaka, Japan). Twenty-five percent of the samples were analyzed in parallel after sampling into tubes containing Paraoxon (Sigma Chemical Co, St Louis, MO) to check for ongoing lipolysis (36). There were no differences in the concentrations of FFAs between samples with and without paraoxon. For glucagon measurements, 0.55 mL aprotinin (Trasylol; Bayer AG, Leverkusen, Germany) was added to chilled, heparin-containing tubes, and the samples were centrifuged (1500 x g, 20 min, 4 °C) before freezing at –80 °C. Glucagon, insulin, C-peptide, proinsulin, leptin, and adiponectin were analyzed by using human-specific RIA kits (Linco Research Co, St Charles, MO). According to the manufacturer, the intra- and interassay CVs vary within the following ranges, respectively: 4.0–6.8% and 7.3–13.5% (glucagon), 2.2–4.4% and 2.9–6.0% (insulin), 3.4–6.4% and 2.4–9.3% (C-peptide), 1.5–6.9% and 1.5–10.1% (proinsulin), 3.4–8.3% and 3.0–6.2% (leptin), and 1.8–6.2% and 6.9–9.3% (adiponectin).

Beta-hydroxybutyrate was measured with a Precision Xtra device (Abbott Laboratories, MediSense Products, Bedford, MA) (37) in thawed serum and EDTA-containing plasma samples.

Statistics and presentation of results
Statistics
Statistical analyses were performed with SPSS version 13.0 (SPSS Inc, Chicago, IL, 2005). Assumptions of normality were checked by visual inspection of normal Q-Q Plots. For the measurements of blood glucose, insulin sensitivity, and insulin secretion, a possible group effect was expected mainly at the last time point (9 wk). Hence, we analyzed these data by using analyses of covariance (ANCOVAs; 38), that is, as regression analyses with the 9-wk measurement as the dependent variable and baseline measurement and group as covariates. The same methods were used for the fatty acid content of the diet and adipose tissue, because these data were sampled only at 2 time points (at baseline and at the end of the intervention). We used repeated-measures analysis of variance (ANOVA) for time and treatment interaction between groups when inspection of the data showed a clear time-dependent effect (eg, indirect calorimetry measurements or fatty acid content of plasma phospholipids). If there was a significant group x time interaction, within-group analyses of the change over time were analyzed by repeated-measures ANOVA separately within each group. Differences between time points within the groups were then compared by using Bonferroni adjustments. If the P value for Mauchly's test of sphericity was ≥0.05, the method assuming sphericity was used; otherwise, the Greenhouse-Geisser method was used.

Power calculations were based on paired t tests and group effects of 12% for insulin sensitivity [measured by hyperinsulinemic clamp (39)] and 20% for insulin secretion [measured by a C-peptide glucagon test (40)] and SDs of 10% and 14%, respectively. A sample size of 12 per group gave a power of 80% and 92%, respectively, with a significance level of 5%.

Presentation of results
Results are given as median values and the variability as the interquartile range (IQR; the distance between the 75th and 25th percentile values) in the text and in the tables as indicated. Median values were preferred for variables where skewness of data was detected or suspected, as in the case of some dietary variables. These data were then analyzed by nonparametric tests (unpaired Mann-Whitney test between groups, paired Wilcoxon's signed-ranks test within groups). Mean values were reported for variables analyzed by ANCOVA or repeated-measures ANOVA. In tables giving means, the variability is expressed as 95% CIs (lower to upper bound). Analyses of log-transformed data gave similar results (not shown). Spearman's correlation coefficients (r) were used to evaluate bivariate correlations. P values < 0.05 (two-sided) were considered significant.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Baseline characteristics
Age, duration of diabetes, blood pressure, body mass index, glycated hemoglobin, fasting serum glucose, and lipids did not differ significantly between groups at baseline (Table 2Go). The intake of fatty fish also did not differ significantly (data not shown), whereas fasting C-peptide concentrations were lower in the corn oil group. Five subjects in the corn oil group and 7 in the fish oil group did not use antidiabetic medication. Metformin alone was used by 5 subjects in the corn oil group and 4 in the fish oil group; metformin combined with glimepiride by 1 and 1 subject, respectively; and glipizide or glimepiride alone by 3 and 0 subjects, respectively. Statins were used by 2 subjects in the corn oil group and 2 in the fish oil group; antihypertensive medication was used by another 2 and 2 subjects, respectively; and estrogen-progesterone was used by 1 subject in the corn oil group and 0 subjects in the fish oil group. The frequency and intensity of physical activity did not differ significantly between groups (see Supplemental Table 1Go in the current issue online at www.ajcn.org).


View this table:
[in this window]
[in a new window]
  		TABLE 2 Baseline variables

 
Self-reported data on diet indicated no significant differences in total energy or nutrient intake (Table 3Go). Energy intake expressed as kcal/kg LBM was 33.6 (6.6) [median (IQR)] in the corn oil group and 33.6 (12.5) in the fish oil group. The relative contribution of energy from different macronutrients did not differ significantly between groups (Table 3Go). In addition, baseline protein intake in the combined group (n = 26) correlated with the amount of protein metabolized, which was calculated from urinary loss of nitrogen (r = 0.605, P < 0.001). This was also found within both study groups.


View this table:
[in this window]
[in a new window]
  		TABLE 3 Daily intake and distribution of energy and nutrients at baseline and during the intervention1

 
There was no significant difference between groups in the percent of energy from SFAs, MUFAs, total n–6 fatty acids, or total n–3 fatty acids (Table 3Go), and there were no baseline differences in specific n–6 and n–3 fatty acids in the diet, in adipose tissue (Table 4Go, 0 wk), or in plasma phospholipids (Table 5Go, 0 wk). A complete fatty acid pattern of diet, adipose tissue, and plasma phospholipids is given in Supplemental Tables 2Go and 3Go in the current issue online at www.ajcn.org.


View this table:
[in this window]
[in a new window]
  		TABLE 4 Fatty acid content of the diet and adipose tissue1

 

View this table:
[in this window]
[in a new window]
  		TABLE 5 Plasma phospholipid n–6 and n–3 fatty acids at baseline (0 wk) and after 1 and 9 wk of intervention1

 
Effects of intervention
Anthropometric variables
Changes in body weight, body fat mass, LBM, and waist circumference during the intervention were minor and did not differ significantly between groups (see Supplemental Table 4Go in the current issue online at www.ajcn.org).

Diet
The intake of oil was close to the target in both groups [17.8 (2.3) mL/d in the corn oil group and 17.6 (2.0) mL/d in the fish oil group]. This corresponds with 8.5 (1.1) g/d of linoleic acid (18:2n–6) in the corn oil group and 5.9 (0.7) g/d of total n–3 fatty acids in the fish oil group. Total n–3 fatty acids included 1.8 (0.2) g eicosapentaenoic acid (EPA; 20:5n–3), 0.3 (0.03) g docosapentaenoic acid (DPA; 22:5n–3), and 3.0 (0.3) g docosahexaenoic acid (DHA; 22:6n–3). The intervention resulted in significant differences between groups in percent of energy from total n–6 fatty acids and n–3 fatty acids (Table 3Go). Because of a higher content of MUFA in the fish oil (Table 1Go), the percent of energy intake from MUFAs differed between groups during the intervention (Table 3Go). Supplements of vitamin C and E amounted to 60 (3) and 51 (7) mg/d, respectively, in the corn oil group and 58 (4) and 52 (6) mg/d, respectively, in the fish oil group, which explains the significant changes from the corresponding baseline intakes (Table 3Go).

Although the oil preparations increased energy intake by 145 (19) and 143 (16) kcal/d in the corn oil and fish oil groups, respectively, there was no significant change in total energy intake during the intervention compared with baseline (Table 3Go), and no significant difference between groups when analyzed by ANCOVA with adjustment for baseline total energy intake (data not shown). Similarly, intakes of protein and carbohydrates were not significantly reduced during the intervention compared with baseline in either group (Table 3Go). The increased intake of total fat was accompanied by an expected reduction in the percent of energy from carbohydrate in both groups, and to a lesser extent from protein (Table 3Go), and the effect did not differ significantly between groups (ANCOVA, data not shown).

n–6 and n–3 Fatty acids in diet, adipose tissue, and plasma phospholipids
According to the ANCOVA with adjustment for baseline dietary intake, the calculated dietary intake of 18:2n–6 increased by 6.4 g/d in the corn oil group, whereas 20:4n–6, 20:5n–3, 22:5n–3, and 22:6n–3 increased by 0.05, 1.84, 0.27, and 3.02 g/d, respectively, in the fish oil group (Table 4Go). At the same time, the dietary ratio of n–6 to n–3 fatty acids was lowered by 6.41 (P < 0.001) in the fish oil group compared with the corn oil group, whereas the ratio of DHA to EPA was not significantly different (difference of –0.11, P = 0.227; Table 4Go.)

Both 20:5n–3 and 22:6n–3 increased in adipose tissue after 9 wk in the fish oil group (Table 4Go). After adjustment for differences in baseline concentrations (ANCOVA), the calculated increases in 20:5n–3 and 22:6n–3 were 0.06 (P = 0.001) and 0.08 mg/100 mg (P = 0.002), respectively. The median increases in DHA and EPA in adipose tissue in the fish oil group were 13% and 18%, respectively (data not shown). Ratios of n–6 to n–3 fatty acids as well as of DHA to EPA decreased during the intervention in the fish oil group (Table 4Go). When adjusted to baseline concentrations and compared with the corn oil group, the decreases were 0.64 (P = 0.030) and 0.41 (P = 0.020), respectively.

As shown in Tables 4Go and 5Go, the dietary increases in 20:5n–3, 22:5n–3, and 22:6n–3 in the fish oil group resulted in a parallel increase in plasma phospholipids already after 1 wk, which increased further after 9 wk. Interestingly, adipose tissue also showed increased concentrations of 20:5n–3 and 22:6n–3 after 9 wk in the fish oil group compared with the corn oil group. The plasma phospholipid ratios of n–6 to n–3 fatty acids and of DHA to EPA were reduced in the fish oil group but not in the corn oil group, with DHA increasing relatively less than EPA (56% compared with 260%, data not shown). The 20:4n–6 concentration decreased in plasma phospholipids in the fish oil group, whereas the reduction was not statistically significant in adipose tissue. In contrast with the effect of dietary changes in n–3 fatty acids on plasma phospholipids and adipose tissue, the increased intake of 18:2n–6 in the corn oil group had no effect on plasma phospholipid and adipose tissue fatty acid concentrations, whereas plasma phospholipid 18:2n–6 decreased in the fish oil group even though dietary intake was unchanged.

Lipoproteins and hormones
The intervention had no significant effects on total cholesterol, LDL cholesterol, HDL cholesterol, triacylglycerol, leptin, adiponectin, glucagon, or insulin, but proinsulin tended to increase in the fish oil group compared with the corn oil group (P = 0.064 by ANOVA; Supplemental Table 4Go in the current issue online at www.ajcn.org).

Blood glucose
Neither the 5 daytime home measurements nor average blood glucose differed significantly between groups at baseline. Average daytime blood glucose concentrations and fasting blood glucose concentrations were significantly higher after 8 wk in the fish oil group than in the corn oil group when adjusted for baseline values (ANCOVA, Table 6Go). The increase was nonlinear with time, reaching a new and higher concentration already after 1 wk (data not shown). Intervention with n–3 fatty acids did not significantly affect glycated hemoglobin (P = 0.093 by ANOVA; Supplemental Table 4Go in the current issue online at www.ajcn.org).


View this table:
[in this window]
[in a new window]
  		TABLE 6 Home blood glucose measurements at 8 wk compared with baseline and of insulin sensitivity and insulin secretion at 9 wk compared with baseline1

 
Ketones
ß-Hydroxybutyrate was below the limit of detection (0.2 mmol/L) in all samples measured, ie, the 10 subjects with the highest fasting glucose concentrations after 9 wk.

Insulin sensitivity as measured by hyperinsulinemic isoglycemic clamps
Fasting insulin concentrations before each clamp were 90 (56), 83 (69), and 69 (69) pmol/L in the corn oil group at baseline, 1 wk, and 9 wk, respectively, and did not differ significantly from the fish oil group, in which the concentrations were 97 (56), 118 (69), and 76 (63) pmol/L, respectively. Hyperinsulinemia was achieved after {approx}15 min, and insulin concentrations during the last 40 min of the clamp were 590 (236), 688 (333), and 674 (292) pmol/L at baseline, 1 wk, and 9 wk in the corn oil group, which were not significantly different from concentrations in the fish oil group [639 (271), 771 (278), and 611 (243) pmol/L, respectively]. Isoglycemia was achieved in both groups at baseline, 1 wk, and 9 wk. The average blood glucose concentrations during the last 40 min of the clamp were 6.6 (2.4), 6.8 (2.4), and 6.9 (2.4) mmol/L in the corn oil group at baseline, 1 wk, and 9 wk, respectively. The corresponding concentrations in the fish oil group were 6.2 (0.8), 6.2 (1.0), and 6.3 (1.5) mmol/L.

Glucose utilization, which is a measure of insulin sensitivity, did not differ significantly between groups at baseline (Table 6Go). When analyzing the effect of the intervention with adjustment for baseline values by ANCOVA, glucose utilization was significantly lowered by 1.56 mg·min–1·kg LBM–1 in the fish oil group compared with the corn oil group at 9 wk.

Neither the corn oil nor the fish oil intervention had any significant effect on fasting FFAs (Supplemental Table 4Go in the current issue online at www.ajcn.org). The clamp markedly reduced total FFAs both at baseline, 1 wk, and 9 wk. At baseline, FFAs decreased from 0.55 (0.21) to 0.06 (0.06) mmol/L in the corn oil group and from 0.67 (0.26) to 0.10 (0.11) mmol/L in the fish oil group. Neither the start (0 min) nor the end (120 min) concentrations, nor the concentration changes during the clamp, differed significantly between groups. At 9 wk however, the decrease in FFA concentration was less marked in the fish oil group than in the corn oil group when analyzed by ANCOVA with adjustment for baseline values (P = 0.005).

Insulin secretion
The stimulated C-peptide response to glucagon (a measure of insulin secretion) did not differ significantly between groups at baseline (Table 6Go). The increased response in the fish oil group at 9 wk was 0.29 nmol/L, which tended to be different from that in the corn oil group (P = 0.078 calculated by ANCOVA with adjustment for baseline values, Table 6Go). Insulin concentrations showed similar trends (data not shown).

Energy metabolism, fasting
The total EPR in the fasting state did not differ significantly between the groups at baseline and did not change during the intervention. Neither did the EPR expressed per kg LBM (Table 7Go). The EPR correlated strongly with LBM (r = 0.794, P < 0.001, n = 26).


View this table:
[in this window]
[in a new window]
  		TABLE 7 Energy metabolism (fasting) measured by indirect calorimetry at baseline (0 wk) and after 1 and 9 wk of intervention1

 
As shown in Table 7Go, there was a significant time and treatment interaction for the nonprotein RQ. In the fish oil group, the nonprotein RQ was increased at 1 wk followed by a decrease at 9 wk. Carbohydrate utilization was thus temporarily increased after 1 wk, after which it was decreased with a concomitant increase in fat utilization.

Energy metabolism during the clamp
Indirect calorimetry was performed in parallel with the insulin and glucose infusion during the clamps some hours after the calorimetry measurements made in the fasted state. Because of practical problems, only 10 subjects in the corn oil group and 9 in the fish oil group successfully completed the calorimetry during all clamps. The differences between groups in terms of preference of energy substrate in the fasting state were accompanied by similar tendencies during the clamp procedures. At baseline, the subjects had a higher nonprotein RQ during the clamp procedures, 0.89 (0.10) compared with 0.86 (0.12) in the fasting state (P = 0.016, n = 19; Wilcoxon's test), which showed that they used more carbohydrate as substrate during the clamp procedure than during fasting. At 9 wk, the nonprotein RQ was 0.81(0.14) during the clamp in the fish oil group compared with 0.92 (0.21) in the corn oil group. However, repeated-measures ANOVA for the whole time-course of energy utilization during the clamp procedures showed no significant group difference (P = 0.357).

Both groups exhibited the same pattern of protein metabolism at baseline, 1 wk, and 9 wk, ie, a smaller amount of protein was metabolized during the clamp than during fasting. This was confirmed by a reduced rate of nitrogen excretion in the urine. For example, during the baseline clamp procedures, the excretion of nitrogen was 12.0 (5.5) compared with 14.8 (6.4) g/24 h in the fasted state (P = 0.002, n = 19; Wilcoxon's test).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
The literature abounds with studies of the effects of dietary supplementation with n–3 fatty acids, and several studies have been performed in subjects with type 2 diabetes. However, the results of intervention studies, especially in populations with type 2 diabetes, are hard to interpret because of the heterogeneity of study populations, possible confounding factors, and limited data on the mechanisms behind the recorded effects. The present study included a comparatively homogeneous, large study population of subjects with type 2 diabetes. Furthermore, the study had a blinded and parallel design and included repeated measurements of metabolic control, insulin sensitivity, insulin secretion, and energy metabolism, which allowed us to assess the time course of effects of n–3 fatty acids on these variables.

n–3 Fatty acids reduce hypertriacylglycerolemia, which may affect insulin sensitivity. Therefore, subjects with hypertriacylglycerolemia were not included in the present study. Triacylglycerol data from other studies, however, eg, the UKPDS (41), indicate that our study population is still representative of a large segment of subjects with type 2 diabetes.

Our results show a moderate but significant deterioration of glycemia in subjects treated with fish oil compared with corn oil. This deterioration was probably not caused by changes in lifestyle, because weight and body composition were stable, and diet did not change during the study period.

The moderate deterioration of metabolic control in the fish oil group was paralleled by lesser glucose utilization than in the corn oil-treated subjects. One might argue that insulin sensitivity was not worsened in the fish oil group but rather improved in the corn oil group. However, improvement in a control group cannot be distinguished from the unspecific effect of trial participation. It is well known that study participation per se usually improves metabolic control in persons with type 2 diabetes.

It is interesting that the FFA-lowering effect of insulin during the clamp procedure was somewhat compromised by a high dose of n–3 fatty acids. Insulin resistance in type 2 diabetes causes a reduced insulin effect on both glucose and FFAs (42, 43). This could indicate common defects in insulin signaling in type 2 diabetes on one hand and effects of n–3 fatty acids on the other. The mechanisms behind these defects remain to be elucidated.

Attenuation of insulin secretion is an alternative cause for deterioration of metabolic control. Our findings rule out this explanation, because C-peptide responses to glucagon were not decreased but tended to be enhanced by fish oil compared with corn oil treatment. Such enhancement can be readily explained as an adaptive response to decreased insulin sensitivity.

Our study showed that n–3 fatty acids time-dependently alter the proportion of carbohydrate and fat utilization as assessed by indirect calorimetry in the fasting state. Notably, these effects did not result in any change in total energy intake, which was stable in all subjects during the intervention. Furthermore, resting energy expenditure was unchanged after both 1 wk and 9 wk of intervention. This finding agrees with the results of other studies (44). In contrast, n–3 fatty acids strikingly affected the ratio between fat and carbohydrate oxidation, effects that were not to any degree seen after corn oil supplementation. Interestingly, the fish oil effects were markedly time-dependent, with reduced fatty acid utilization after 7 d that changed to increased fatty acid utilization after 9 wk. The calorimetry data obtained during the clamp procedure were compatible with the results obtained in the fasting state. Thus, there was a relative, although not significant, increase in fat utilization in the fish oil group during the clamp procedures performed at the end of the study.

Our findings of increased fat utilization after 9 wk agree with results found in nondiabetic persons (45). That previous study suggested increased ketogenesis as a possible explanation. However, we did not find that n–3 fatty acids elevated ketones. Our observation that the increased fatty acid utilization is time dependent indicates an induction of enzymes essential for fatty acid mitochondrial or peroxisomal oxidation.

Puhakainen et al (44) found that n–3 fatty acids increased gluconeogenesis from glycerol. Such an effect could theoretically induce insulin resistance. However, the application of these results to our study is doubtful, because Puhakainen et al did not find deterioration of metabolic control or increased fat oxidation.

It may seem paradoxical that the FFA-lowering effect of insulin during the clamps was reduced in the fish oil group; yet fat oxidation increased in proportion to carbohydrate oxidation. However, plasma FFA concentrations generally reflect the rate of generation of fatty acids rather than their rate of removal (46). Thus, ongoing lipolysis during the clamp could maintain concentrations of circulating FFA.

We used a high dose of n–3 fatty acids because this would facilitate the detection of untoward effects in a time-course perspective. Furthermore, should we not find any such effects, this would confirm that n–3 fatty acids could be recommended to subjects with type 2 diabetes without restrictions on dosage. Clearly, our results do not justify recommending high-dose n–3 fatty acid dietary supplementation. On the other hand, we note that the untoward effects were moderate. Given a linear dose-response relation (which seems plausible but not proven), intake of 1 g daily, which is commonly recommended in Norway, should have negligible untoward effects on metabolic control in subjects with type 2 diabetes. Thus, our findings do not rule out the use of n–3 fatty acids as a preventive measure in subjects at risk of cardiovascular disease.

Other studies used more EPA than DHA in their supplements (most often 1.8 g EPA and 1.2 g DHA). In the present study, the amount of DHA the subjects ingested was 2.5-fold that of EPA. These proportions reflect those seen in the cod liver oil made in Norway. This is different from the proportions EPA to DHA in most commercial n–3 supplements, because these supplements are based on fish oils (anchovies and sardines) containing 18% EPA and 12% DHA.

Our phospholipid and adipose tissue data suggest that EPA and DHA are handled differently. Calculated as a percent, more EPA than DHA was incorporated into plasma phospholipids and adipose tissue. This agrees with previous results (47) showing that supplementation with DHA increases both DHA and EPA in serum phospholipids, indicating DHA retroconversion to EPA.

Corn oil was chosen as a placebo because it was available with the same flavor and taste as the fish oil and contained fatty acids already commonly used in the Norwegian population. We found that linoleic acid did not increase in plasma phospholipids in the control group despite the marked increase in diet, which indicates that plasma phospholipids were saturated with linoleic acid. In the fish oil group, the expected increase in plasma phospholipid n–3 fatty acids and decrease in n–6 fatty acids were clearly seen.

The KANWU study showed that substituting dietary SFAs with MUFAs increases insulin sensitivity only in healthy subjects consuming <37% of energy from fat (48). At higher fat intakes, the authors found no difference between SFAs and MUFAs in effects on insulin sensitivity. In our study, the subjects in the corn oil group and fish oil group had initial fat intakes of 31.2% and 31.9% of energy, respectively, which increased during the intervention to 39.1% and 39.4% of energy, respectively. It thus seems that the effect of n–3 fatty acid supplementation on insulin sensitivity is not limited to a fat intake below 37% of energy in subjects with type 2 diabetes.

To conclude, a high intake of fish oil compared with corn oil has moderate untoward effects on glycemia and insulin sensitivity but not on insulin secretion. Furthermore, a high intake of n–3 fatty acids time-dependently alters the relative proportions of fat and carbohydrate oxidation. Because of the reduced insulin sensitivity in subjects with type 2 diabetes without hypertriacylglycerolemia, these persons should be discouraged from supplementing their diet with large amounts of n–3 fatty acids. Our data, however, do not imply that a moderate intake of marine n–3 fatty acids will have a detrimental effect in such patients.


   ACKNOWLEDGMENTS
 
We are grateful to the subjects who participated in the study. We thank the nurses Anne Marit Aukan and Sissel Johnsen for competent assistance during the clamp procedures, Inge Thyve and Christian Hermstad for support in monitoring the Deltatrac II metabolic monitor and YSI glucose analyzer, Ingrid Hals Jørgensen for assistance with the hormone analyses, and the medical and secretarial staff at the Division of Endocrinology and the staff at the Department of Medical Biochemistry, St Olavs Hospital, for help in various ways.

VG initiated and designed the study together with ILM. VG took part in data analyses and manuscript writing. ILM was responsible for and participated in all aspects of the study, including planning and designing; performing the clamp procedures and hormone and calorimetric analyses; and the major part of manuscript writing. SL and ILM carried out the statistical analyses. KSB was responsible for evaluating many of the laboratory analyses and advising on study design and participating in manuscript writing. MRB contributed in designing the clamp procedures, in evaluating the results, and in manuscript writing. None of the authors had a conflict of interest.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Rustan AC, Nenseter MS, Drevon CA. Omega-3 and omega-6 fatty acids in the insulin resistance syndrome. Lipid and lipoprotein metabolism and atherosclerosis. Ann N Y Acad Sci 1997;827:310–26.[Abstract/Free Full Text]
  2. Simopoulos AP. Omega-3 fatty acids in health and disease and in growth and development. Am J Clin Nutr 1991;54:438–63.[Abstract/Free Full Text]
  3. Sosial- og helsedirektoratet. [The Norwegian Directorate for Health and Social Affairs.] Mat og mosjon i hverdagen. [Daily food and exercise]. Oslo, Norway: Sosial- og helsedirektoratet, 2004 (in Norwegian).
  4. Statens ernæringsråd. [The National Nutrition Council.] Helse&Kost—om hvordan matvanene påvirker helsen. [Health&Food—how dietary habits influence health.] Oslo, Norway: Statens ernæringsråd, 1994 (in Norwegian).
  5. Norges Diabetesforbund. [The Norwegian Diabetes Association.] Mat og diabetes. [Food and diabetes.] Oslo, Norway: Norges Diabetesforbund, 2004 (In Norwegian).
  6. Storlien LH, Baur LA, Kriketos AD, et al. Dietary fats and insulin action. Diabetologia 1996;39:621–31.[Medline]
  7. Montori VM, Farmer A, Wollan PC, Dinneen SF. Fish oil supplementation in type 2 diabetes: a quantitative systematic review. Diabetes Care 2000;23:1407–15.[Abstract/Free Full Text]
  8. Rivellese AA, Lilli S. Quality of dietary fatty acids, insulin sensitivity and type 2 diabetes. Biomed Pharmacother 2003;57:84–7.[Medline]
  9. Vessby B, Boberg M. Dietary supplementation with n–3 fatty acids may impair glucose homeostasis in patients with non-insulin-dependent diabetes mellitus. J Intern Med 1990;228:165–71.[Medline]
  10. Boberg M, Pollare T, Siegbahn A, Vessby B. Supplementation with n–3 fatty acids reduces triglycerides but increases PAI-1 in non-insulin-dependent diabetes mellitus. Eur J Clin Invest 1992;22:645–50.[Medline]
  11. Schectman G, Kaul S, Kissebah AH. Effect of fish oil concentrate on lipoprotein composition in NIDDM. Diabetes 1988;37:1567–73.[Abstract]
  12. Borkman M, Chisholm DJ, Furler SM, et al. Effects of fish oil supplementation on glucose and lipid metabolism in NIDDM. Diabetes 1989;38:1314–9.[Abstract]
  13. Axelrod L, Camuso J, Williams E, Kleinman K, Briones E, Schoenfeld D. Effects of a small quantity of omega-3 fatty acids on cardiovascular risk factors in NIDDM. A randomized, prospective, double-blind, controlled study. Diabetes Care 1994;17:37–44.
  14. Hendra TJ, Britton ME, Roper DR, et al. Effects of fish oil supplements in NIDDM subjects. Controlled study. Diabetes Care 1990;13:821–9.[Abstract]
  15. Sirtori CR, Paoletti R, Mancini M, et al. N–3 fatty acids do not lead to an increased diabetic risk in patients with hyperlipidemia and abnormal glucose tolerance. Italian Fish Oil Multicenter Study. Am J Clin Nutr 1997;65:1874–81.[Abstract/Free Full Text]
  16. Connor WE, Prince MJ, Ullmann D, et al. The hypotriglyceridemic effect of fish oil in adult-onset diabetes without adverse glucose control. Ann N Y Acad Sci 1993;683:337–40.[Abstract]
  17. Pelikanova T, Kohout M, Valek J, Kazdova L, Base J. Metabolic effects of omega-3 fatty acids in type 2 (non-insulin-dependent) diabetic patients. Ann N Y Acad Sci 1993;683:272–8.[Abstract]
  18. McManus RM, Jumpson J, Finegood DT, Clandinin MT, Ryan EA. A comparison of the effects of n–3 fatty acids from linseed oil and fish oil in well-controlled type II diabetes. Diabetes Care 1996;19:463–7.[Abstract]
  19. Annuzzi G, Rivellese A, Capaldo B, et al. A controlled study on the effects of n–3 fatty acids on lipid and glucose metabolism in non-insulin-dependent diabetic patients. Atherosclerosis 1991;87:65–73.[Medline]
  20. Morgan WA, Raskin P, Rosenstock J. A comparison of fish oil or corn oil supplements in hyperlipidemic subjects with NIDDM. Diabetes Care 1995;18:83–6.[Abstract]
  21. Westerveld HT, de Graaf JC, van Breugel HH, et al. Effects of low-dose EPA-E on glycemic control, lipid profile, lipoprotein(a), platelet aggregation, viscosity, and platelet and vessel wall interaction in NIDDM. Diabetes Care 1993;16:683–8.[Abstract]
  22. Luo J, Rizkalla SW, Vidal H, et al. Moderate intake of n–3 fatty acids for 2 months has no detrimental effect on glucose metabolism and could ameliorate the lipid profile in type 2 diabetic men. Results of a controlled study. Diabetes Care 1998;21:717–24.[Abstract]
  23. Popp-Snijders C, Blonk MC. Omega-3 fatty acids in adipose tissue of obese patients with non-insulin-dependent diabetes mellitus reflect long-term dietary intake of eicosapentaenoic and docosahexaenoic acid. Am J Clin Nutr 1995;61:360–5.[Abstract/Free Full Text]
  24. Andersen LF, Solvoll K, Drevon CA. Very-long-chain n–3 fatty acids as biomarkers for intake of fish and n–3 fatty acid concentrates. Am J Clin Nutr 1996;64:305–11.[Abstract/Free Full Text]
  25. Pocock SJ, Simon R. Sequential treatment assignment with balancing for prognostic factors in the controlled clinical trial. Biometrics 1975;31:103–15.[Medline]
  26. Jensen CV. A computer program for randomizing patients with near-even distribution of important parameters. Comput Biomed Res 1991;24:429–34.[Medline]
  27. DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol 1979;237:E214–23.
  28. Ferrannini E. The theoretical bases of indirect calorimetry: a review. Metabolism 1988;37:287–301.[Medline]
  29. Takala J, Meriläinen P. Handbook of gas exchange and indirect calorimetry. Helsinki, Finland: Datex-Engstrom Division, Instrumentarium Corp, 2001.
  30. Rimestad AH, Blaker B, Flåten A-M, Nordbotten A. Den store matvaretabellen. [The Norwegian food table.] Oslo, Norway: Universitetsforlaget, 1995 (in Norwegian).
  31. Bonaa KH, Bjerve KS, Straume B, Gram IT, Thelle D. Effect of eicosapentaenoic and docosahexaenoic acids on blood pressure in hypertension. A population-based intervention trial from the Tromso study. N Engl J Med 1990;322:795–801.[Abstract]
  32. Beynen AC, Katan MB. Rapid sampling and long-term storage of subcutaneous adipose-tissue biopsies for determination of fatty acid composition. Am J Clin Nutr 1985;42:317–22.[Abstract/Free Full Text]
  33. Pedersen JI, Ringstad J, Almendingen K, Haugen TS, Stensvold I, Thelle DS. Adipose tissue fatty acids and risk of myocardial infarction—a case-control study. Eur J Clin Nutr 2001;54:618–25.
  34. Viga A, Grahl-Nielsen O. Genotypic and phenotypic fatty-acid composition in the tissue of salmon (Salmo salar). Comp Biochem Physiol 1990;96B:721–7.
  35. 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;18:499–502.[Abstract]
  36. Zambon A, Hashimoto SI, Brunzell JD. Analysis of techniques to obtain plasma for measurement of levels of free fatty acids. J Lipid Res 1993;34:1021–8.[Abstract]
  37. Byrne HA, Tieszen KL, Hollis S, Dornan TL, New JP. Evaluation of an electrochemical sensor for measuring blood ketones. Diabetes Care 2000;23:500–3.[Abstract]
  38. Vickers AJ, Altman DG. Statistics notes: analysing controlled trials with baseline and follow up measurements. BMJ 2001;323:1123–4.[Free Full Text]
  39. Ferrannini E, Mari A. How to measure insulin sensitivity. J Hypertens 1998;16:895–906.[Medline]
  40. Gottsater A, Landin-Olsson M, Fernlund P, Gullberg B, Lernmark A, Sundkvist G. Pancreatic beta-cell function evaluated by intravenous glucose and glucagon stimulation. A comparison between insulin and C-peptide to measure insulin secretion. Scand J Clin Lab Invest 1992;52:631–9.[Medline]
  41. Davis TM, Wright AD, Mehta ZM, et al. Islet autoantibodies in clinically diagnosed type 2 diabetes: prevalence and relationship with metabolic control (UKPDS 70). Diabetologia 2005;48:695–702.[Medline]
  42. Reaven GM. Banting lecture 1988. Role of insulin resistance in human disease. Diabetes 1988;37:1595–607.
  43. Groop LC, Bonadonna RC, DelPrato S, et al. Glucose and free fatty acid metabolism in non-insulin-dependent diabetes mellitus. Evidence for multiple sites of insulin resistance. J Clin Invest 1989;84:205–13.[Medline]
  44. Puhakainen I, Ahola I, Yki-Jarvinen H. Dietary supplementation with n–3 fatty acids increases gluconeogenesis from glycerol but not hepatic glucose production in patients with non-insulin-dependent diabetes mellitus. Am J Clin Nutr 1995;61:121–6.[Abstract/Free Full Text]
  45. Delarue J, Couet C, Cohen R, Brechot JF, Antoine JM, Lamisse F. Effects of fish oil on metabolic responses to oral fructose and glucose loads in healthy humans. Am J Physiol 1996;270:E353–62.
  46. Bonadonna RC, Groop LC, Zych K, Shank M, DeFronzo RA. Dose-dependent effect of insulin on plasma free fatty acid turnover and oxidation in humans. Am J Physiol 1990;259:E736–50.
  47. Grimsgaard S, Bonaa KH, Hansen JB, Nordoy A. Highly purified eicosapentaenoic acid and docosahexaenoic acid in humans have similar triacylglycerol-lowering effects but divergent effects on serum fatty acids. Am J Clin Nutr 1997;66:649–59.[Abstract/Free Full Text]
  48. Vessby B, Unsitupa M, Hermansen K, et al. Substituting dietary saturated for monounsaturated fat impairs insulin sensitivity in healthy men and women: the KANWU Study. Diabetologia 2001;44:312–9.[Medline]
Received for publication September 5, 2005. Accepted for publication April 20, 2006.




This article has been cited by other articles:


Home page
 Diabetes CareHome page
V. Hainer, H. Toplak, and A. Mitrakou
Treatment Modalities of Obesity: What fits whom?
Diabetes Care, February 1, 2008; 31(Supplement_2): S269 - S277.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Clin. Nutr.Home page
M. Kabir, G. Skurnik, N. Naour, V. Pechtner, E. Meugnier, S. Rome, A. Quignard-Boulange, H. Vidal, G. Slama, K. Clement, et al.
Treatment for 2 mo with n 3 polyunsaturated fatty acids reduces adiposity and some atherogenic factors but does not improve insulin sensitivity in women with type 2 diabetes: a randomized controlled study
Am. J. Clinical Nutrition, December 1, 2007; 86(6): 1670 - 1679.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Tables
Right arrow An erratum has been published
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mostad, I. L
Right arrow Articles by Grill, V.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mostad, I. L
Right arrow Articles by Grill, V.
Agricola
Right arrow Articles by Mostad,