Effect of a Low-Fat Versus a Low-Gycemic-Load Diet on Inflammatory Biomarker and Adipokine Concentrations

Abstract

Background:

Low-grade inflammation is linked to metabolic syndrome and obesity. We studied the effects of weight loss and dietary composition on serum concentrations of biomarkers of inflammation and adipokines.

Methods:

Men and women (n=181) aged 30–65 years with a body mass index (BMI) of 28–40 kg/m² (28–35 kg/m² for women) and one or more components of metabolic syndrome were randomized to follow one of two hypocaloric diets, a low-fat or low-glycemic-load diet for 3 months. Blood samples were taken pre- and postintervention. Serum concentrations of interleukin-6 (IL-6), tumor necrosis factorα (TNF-α), plasminogen activator inhibitor-1 (PAI-1), monocyte chemoattractant protein-1 (MCP-1), and adipokines (leptin, resistin, and adiponectin) were analyzed using multiplexed microsphere immunoassays.

Results:

Weight loss was not different in the low-fat (4.4%±3.8%) and low-glycemic-load (4.9%±3.2%) groups. Concentrations of IL-6, TNF-α, PAI-1, and leptin were significantly reduced in both dietary groups with no between-group differences, whereas MCP-1 and adiponectin concentrations did not change. Subjects with full metabolic syndrome (three or more components; n=109) experienced greater weight loss than subjects (n=72) with one to two components and greater reduction in leptin [7.08 (95% confidence interval 5.19, 8.97) vs. 3.46 (0.91, 6.00) ng/mL; p=0.02] and a tendency to greater reduction in TNF-α (1.00 [0.60, 1.44] vs 0.40 [0.02, 0.78] pg/mL; p=0.05).

Conclusion:

Hypocaloric diets improved inflammatory biomarkers and adipokines independently of dietary composition. The response tended to be greater in subjects with three or more components of metabolic syndrome than their counterparts with one to two components. ClinicalTrials.gov Identifier: NCT00230919

Introduction

A dipose tissue actively produces adipokines and inflammatory markers that may be markers of susceptibility to cardiovascular disease (CVD) and diabetes or subclinical atherosclerosis.¹ Such biomarkers have been examined in regard to their ability to predict CVD^2,3 and to assess the effect of weight loss or other preventive measures on CVD risk.^4,5 Weight loss clearly reverses the proinflammatory profile of the obese state,^4,5 and dietary change and physical activity remain cornerstones of risk reduction regardless of the mechanisms involved. Achieving long-term behavioral adherence in the presence of strong biological imperatives, like low leptin levels, that drive overeating and energy conservation in response to weight loss has proven to be difficult.⁶

No diet has proven much higher effectiveness than others in the long term⁷; however, recent trials have suggested that a low-carbohydrate diet is a feasible alternative to low-fat diets and may have favorable metabolic effects, particularly in subjects with insulin resistance.^8,9 The diets have been compared in regard to their effects on adipokines, but studies have reported contradictory findings,^10,11 and it remains unclear whether diet modulates obesity biomarkers beyond weight loss.¹² In a recent randomized controlled clinical trial, we found that a low-fat diet caused more reduction in body weight and waist circumference than a low-glycemic-load diet in subjects with only one to two criteria of metabolic syndrome, but the low-glycemic-load diet was as effective as the low-fat diet in those with three or more criteria.¹³ This led us to the question of whether benefits of the diets differ between these groups. In the current study, we compare the two diets on markers of inflammation [interleukin 6 (IL-6), monocyte chemoattractant protein-1 (MCP-1), plasminogen activator inhibitor-1 (PAI-1), tumor necrosis factorα (TNF-α, and adipokines (leptin, resistin, and adiponectin) in subjects with one to two versus three or more components of metabolic syndrome to clarify whether metabolic characteristics modify biomarker responses to the diet.

Subjects and Methods

The study was performed at the Department of Preventive Cardiology, Oslo University Hospital, Ullevål, Norway as described previously.¹³ In brief, we included 202 men and women aged 30–65 with a body mass index (BMI) between 28 and 40 kg/m² for men (28 and 35 kg/m² for women) and one or more metabolic syndrome criteria as defined by the National Cholesterol Education Program.¹⁴ Men had a wider range of BMI inclusion criteria because of their high risk and tendency to be underrepresented in studies. Potential subjects with symptomatic cardiovascular disease and individuals taking lipid-lowering or antidiabetic medication were excluded. The regional ethics committee evaluated the protocol. All participants signed a written informed consent form. Of the total, 181 (90%) had complete inflammatory and adipokine biomarkers at 3 months.

Subjects were randomized to follow one of two mildly hypocaloric diets (i.e., −500 kcal/day), a low-fat diet (<30% fat, ≈15% protein, 55%–60% carbohydrates), or a low-glycemic-load diet (35%–40% fat, 25%–30% protein, 30%–35% carbohydrates). Dieticians met with participants at regularly scheduled clinic visits to ensure compliance to the diets. For the low-fat diet, low-fat foods and carbohydrates high in fiber were recommended. Subjects in the low-glycemic-load group were asked to fill one-third of the plate with a primary protein source at all meals, whereas the other two-thirds was to consist of vegetables, salads, and/or legumes for two daily meals. At the third meal, the remaining two-thirds was divided between bread, preferably pumpernickel or other low-glycemic-index breads or basmati rice, whole wheat pasta, and vegetables. Recipes were made available for both groups. At 3 months a 7-day dietary questionnaire was obtained, coded by the Institute of Nutrition Research at the University of Oslo.

Laboratory Analyses

Blood samples were taken at baseline and at 3 months' follow-up after subjects had fasted for at least 10 h. Multiplexed microsphere immunoassays (Lincoplex™, Linco) and a Bio-Plex™ 200 instrument (BioRad) were used to measure serum concentrations of IL-6, TNF-α, total PAI-1, MCP-1, leptin, resistin, and adiponectin in duplicate samples, with the following mean coefficients of variation: 9.1%, 6.1%, 6.6%, 4.7%, 4.5%, 10.5%, and 6.7%, respectively.

Statistical analyses

The distributions of all variables were analyzed with the Shapiro–Wilks test and several were not normally distributed. Log transformation of skewed variables led to similar results to those without transformation; thus, we chose to present results of nontransformed analyses. In a generalized linear model two-way repeated measurements analysis of variance, we tested the interaction between time and diet (see Table 2, below) or number of metabolic syndrome components (see Table 4, below) using the Wilks lambda statistic; 95% confidence intervals were calculated for changes. A two-tailed P<0.05 between groups was considered statistically significant. Analyses were performed with the SPSS 15.00 software program (SPSS Inc. Chicago).

Results

Baseline characteristics of subjects (n=88) assigned to the low-fat diet were not different from those of subjects (n=93) assigned to the low-glycemic-load diet [age, mean±standard deviation (SD), 49.8±8.1 years vs. 50.3±9.4 years; BMI, 33.0±2.7 kg/m² vs. 33.3±3.1 kg/m²; gender, 60.2% female vs. 55.9% female, respectively). Dietary characteristics in the groups are shown in Table 1 for subjects with complete data (n=80 in the low-fat group and n=78 in the low-glycemic-load group). Weight loss (mean±SD) after 3 months did not differ in the low-fat diet group (4.3±4.0 kg; 4.4%±3.8%) compared to the low-glycemic-load diet group (4.9±3.4 kg; 4.9%±3.2%). Serum concentrations of IL-6, PAI-1, TNF-α, and leptin were reduced in both dietary groups; concentrations of resistin were only reduced in the low-fat group (Table 2). MCP-1 and adiponectin concentrations did not change. Changes in inflammatory biomarkers and adipokines did not differ between the dietary groups (Table 2). Change in body weight was correlated with change in leptin concentration (Spearman r=0.358; P<0.001) but not with changes in other adipokines or inflammatory biomarkers (data not shown).

Table 1.

Dietary Characteristics of the Low-Fat and Low-Glycemic-Load Groups

Parameter	Low fat diet group (n=80)	Low glycemic load diet group (n=78)	P value between groups
Energy (kJ/day)	7562±1899	8021±2131	0.16
Fat (% energy)	32.0±5.6%	36.8±6.0%	<0.001
Protein (% energy)	18.6±2.9%	21.2±4.1%	<0.001
Carbohydrates (% energy)	45.2±5.8%	37.2±7.1%	<0.001
Added sugar (% energy)	5.5±2.9%	5.2±3.1%	0.34
Alcohol (% energy)	4.1±3.9%	4.7±4.6%	0.42
Fiber (grams/day)	20.3±10.6	22.4±10.7	0.21

Mean±standard deviation (SD) values are shown.

Table 2.

Concentrations of Inflammatory Biomarkers and Adipokines at Baseline and at 3 Months According to Dietary Group

	Low-fat diet group (n=88)			Low-glycemic-load diet group (n=93)
Parameter	Baseline^a	3 months^a	Change^b	Baseline^a	3 months^a	Change^b	P between diet groups^c
IL-6 (pg/mL)	9.74±9.46	8.55±7.88	1.13 (0.21, 2.05)	11.43±15.39	9.68±12.24	1.67 (0.69, 2.65)	0.43
MCP-1 (pg/mL)	324.02±160.66	305.53±117.29	18.50 (−3.41, 40.41)	305.73±115.39	301.76±126.63	4.14 (−8.00, 16.27)	0.24
PAI-1 (ng/mL)	42.37±13.07	38.40±10.60	3.96 (2.21, 5.72)	42.40±11.46	37.92±9.44	4.47 (2.73, 6.21)	0.68
TNF-α (pg/mL)	7.86±5.30	6.99±4.09	0.87 (0.39, 1.35)	7.40±5.76	6.74±4.78	0.68 (0.32, 1.04)	0.48
Leptin (ng/mL)	29.34±17.26	23.82±16.76	5.52 (3.03, 8.02)	26.57±15.27	20.82±12.75	5.75 (3.88, 7.62)	0.88
Adiponectin (μg/mL)	18.88±10.46	18.17±9.70	−0.71 (−1.71, 0.28)	17.60±10.36	18.01±10.33	0.41 (−0.50, 1.31)	0.10
Resistin (ng/mL)	16.15±8.47	14.77±7.02	1.38 (0.18, 2.58)	14.40±5.97	14.43±6.33	−0.04 (−0.91, 0.83)	0.06

Mean±standard deviation (SD) values are shown.

Difference between baseline and 3 months showing mean and 95% confidence intervals.

General linear model with repeated measurements was used to test values between diet groups.

IL-6, interleukin-6; MCP-1, monocyte chemoattractant protein-1; PAI-1, plasminogen activator inhibitor-1; TNF-α, tumor necrosis factor-α

Subjects were stratified according to the number of components of metabolic syndrome. Subjects with three or more components (metabolic syndrome) were somewhat younger and lost more weight during the intervention than their counterparts with one to two components (Table 3). Baseline concentrations of IL-6 and resistin were higher and the adiponectin concentration was lower in subjects with metabolic syndrome than in those with fewer components (baseline values shown in Table 4).

Table 3.

Baseline Characteristics and Changes After 3 Months for Subjects with Three or More Components of Metabolic Syndrome or with One to Two Components

	Three or more metabolic syndrome components	One to two metabolic syndrome components
Number (%)	109	72
Age (years)	48.5±8.6	52.3±8.6^a
Proportion of females	55.0%	62.5%
Weight (kg)	101.6±16.4	96.7±13.8
BMI (kg/m²)	33.3±3.0	32.8±2.9
Waist circumference (cm)	110.9±9.7	109.1±10.6
Hip circumference (cm)	113.0±6.3	112.8±6.6
Systolic blood pressure (mmHg)	130±14	128±15
Diastolic blood pressure (mmHg)	92±8.6	90±10
Proportion on low fat/low glycemic load diet	45.0%/55.0%	54.2%/45.8%
Weight loss (kg)	5.1±3.3	3.9±4.2^a
Change in waist circumference (cm)	4.7±0.5	4.7±0.5

Mean±standard deviation (SD) or percentages are shown.

P value=0.04.

BMI, body mass index.

Table 4.

Inflammatory Biomarker and Adipokine Concentrations and Changes at Baseline and 3 Months in Subjects with Three or More Components of Metabolic Syndrome or One to Two Components

Parameter	n	Baseline^a	3 months^a	Change^b	P value between groups^c
IL-6 (pg/mL)
≥3 MS components	108	12.84±15.06	10.94±11.59	1.83 (0.85, 2.80)
1–2 MS components	70	7.15±7.14	6.34±7.24	0.76 (−0.04, 1.56)	0.12
MCP-1 (pg/mL)
≥3 MS components	109	310.06±149.67	297.00±126.26	13.06 (−5.19, 31.51)
1–2 MS components	72	321.53±122.24	313.58±115.01	7.95 (−5.78, 21.90)	0.68
PAI-1 (ng/mL)
≥3 MS components	109	43.52±11.79	38.93±10.09	4.59 (3.05, 6.20)
1–2 MS components	72	40.66±12.78	36.97±9.80	3.68 (1.61, 5.60)	0.47
TNF-α (pg/mL)
≥3 MS components	109	8.35±6.58	7.35±5.18	1.00 (0.60, 1.44)
1–2 MS components	72	6.52±3.10	6.13±2.88	0.40 (0.02, 0.78)	0.05
Leptin ng/mL
≥3 MS components	109	27.88±15.14	20.80±12.76	7.08 (5.19, 8.97)
1–2 MS components	72	27.96±17.98	24.50±17.46	3.46 (0.91, 6.00)	0.02
Adiponectin (μg/mL
≥3 MS components	109	16.04±9.19	16.08±9.22	0.04 (−0.77, 0.86)
1–2 MS components	72	21.53±11.29	21.12±10.43	−0.41 (−1.5, 0.77)	0.51
Resistin (ng/mL)
≥3 MS components	109	16.24±8.06	15.51±6.90	0.73 (0.29, 1.75)
1–2 MS components	72	13.76±5.80	13.22±6.07	0.54 (−0.51, 1.58)	0.80

Mean±standard deviation (SD) values are shown.

Difference between baseline and 3 months showing mean and 95% confidence intervals.

General linear model with repeated measurements was used to test values between groups.

IL-6, interleukin-6; MCP-1, monocyte chemoattractant protein-1; PAI-1, plasminogen activator inhibitor-1; TNF-α, tumor necrosis factor-α.

Table 4 shows changes in inflammatory markers and adipokines between baseline and 3 months in subjects with metabolic syndrome and subjects with one to two components. A tendency to reduction in TNF-α and reductions in leptin were greater in subjects with metabolic syndrome, whereas changes in other variables did not differ between the groups.

Discussion

Mean weight loss of <5% regardless of diet was associated with reduced concentrations of several inflammatory biomarkers and adipokines. In the subgroup of participants with metabolic syndrome, IL-6 and resistin concentrations were higher and the adiponectin concentration was lower than in the subgroup with one to two components, indicating a more pronounced inflammatory state. This group experienced a tendency to a greater reduction in TNF-α following weight loss (both diets combined) and a greater reduction in leptin. These findings suggest mechanisms whereby individuals with metabolic syndrome and insulin resistance may lower risk of type 2 diabetes and CVD with quite modest weight loss. Although low carbohydrate diets may have advantages for subjects with insulin resistance,^8,9,13 we were unable to show advantages of these diets in regard to inflammatory biomarkers and adipokine concentrations in such subjects. We did not find any correlation between reported intake of carbohydrates and changes in the analyzed biomarkers (data not shown).

A recent review found contradictory effects of modest mean weight losses of <5% on leptin concentrations.⁵ De Luis et al.¹⁰ observed decreases in leptin with a similar amount of weight loss as the current study in obese subjects not selected for metabolic syndrome. The same authors reported a greater decrease in leptin with a low-carbohydrate versus a low-fat diet, in contrast to our findings, but consistent with others.¹⁵ Diets may vary in degree of restriction of carbohydrates and not be directly comparable for a variety of reasons. Animal studies have shown improvement in leptin concentrations when switching from high-fat diets to high carbohydrates,¹⁶ but may have little relevance to studies in free-living humans. The greater reduction in leptin in the group with full metabolic syndrome is congruent with the greater weight loss in that group.

Although adiponectin is secreted from adipose tissue, plasma concentrations are inversely correlated to body fat or weight and would be expected to increase with weight loss. We found no significant change in adiponectin in the present study, even when we examined the subgroup who lost >5% of their body weight (data not shown). Some studies suggest that a weight loss of >10% is necessary to increase adiponectin concentrations,^4,5,17 or longer follow-up may be needed. Data from the STEDMAN Project support this notion by showing that adiponectin concentration was stable during weight loss, but increased in the follow-up period despite some regain of weight.¹⁸ Lifestyle programs that also involve exercise training may enhance increases in adiponectin concentrations with even modest weight loss.¹⁹

Consistent with the pathophysiological pathways connecting concentrations of resistin to obesity, insulin resistance, and inflammation,²⁰ and the results from a Finnish health survey,²¹ we noted higher baseline concentrations of resistin in the subgroup with metabolic syndrome compared with those with fewer components. Furthermore, resistin and inflammatory biomarker concentrations were positively correlated (data not shown) confirming earlier observations.^21,22 While we observed a reduction in resistin only in the low-fat group in the within-group analysis, the between-diet P value observed was not statistically significant. Earlier studies are contradictory in regard to the effect of modest weight loss on resistin concentrations.⁵

The inflammatory biomarkers, IL-6, TNF-α, and PAI-1 were reduced in both dietary groups with no between-group differences. The reduction in IL-6 is in contrast to the recent review that found that a weight loss of >8% is necessary to reduce this parameter.⁵ However, our subjects were selected for metabolic syndrome features and thereby exhibit a higher state of underlying inflammation and are more likely to show effects of weight reduction than unselected obese subjects. Likewise decreases in concentrations of TNF-α did not differ between diet groups but were greater in the subgroup with metabolic syndrome. In comparison, Seshardi found no significant reduction in TNF-α within dietary groups, but a significant difference between diets as TNF-α increased in conventional (low-fat) dieters.¹⁵ Others have reported no change,^10,23,24 decreases regardless of diet,²⁵ and decreases after about a 10% weight reduction and maintenance period.²⁶

The reduction in PAI-1 was consistent with observations in earlier studies among overweight and obese men and women with metabolic syndrome factors following different diets, although the weight reduction in these studies was of somewhat higher magnitude.^27
–29 Our observation of no significant change in MCP-1 in the dietary groups is in line with the metaanalysis that found no effect of 4%–9% weight loss on MCP-1 concentrations.⁴ Greater weight loss >10% seems to reduce MCP-1.^24,30 A short-term more extreme high-fat, low-carbohydrate diet resulted in 3%–4% weight reduction and a significant reduction in MCP-1.²⁶

In this study, we measured only body weight and waist circumference, and thus lack data concerning fat mass or visceral fat mass. The planned diet goals for each macronutrient were not reached, but the achieved differences in intake between diets were statistically significant.¹³ Our study is a relatively large study and included both men and women in a free-living situation receiving moderate dietary counseling. We measured a wide range of inflammatory markers.

In conclusion, with mean weight reductions of <5% on low-fat or low-carbohydrate diets, overweight and obese men and women with features of metabolic syndrome improved inflammatory biomarkers and adipokines. These advantages seem to be more marked in subjects with three or more components of metabolic syndrome, underscoring the potential health benefits of even a moderate weight loss in these patients.

Footnotes

Acknowledgments

The study was supported by a grant from the Norwegian National Research Council. We thank Hege Thorsrud and Kari Sygnestveit for dietary advice and follow-up, and Lise Bergengen, Lisa Flakk, and Ragnhild Kleve for blood sampling and clinical measurements.

Author Disclosure Statement

The authors declare that no competing interests exist.

References

Gustafson

, Hammarstedt

, Andersson

et al. Inflamed adipose tissue: A culprit underlying the metabolic syndrome and atherosclerosis. Arterioscler Thromb Vasc Biol, 2007; 27:2276–2283.

Berg

, Scherer

PE.

Adipose tissue, inflammation, and cardiovascular disease. Circ Res, 2005; 96:939–949.

Hansson

. Inflammation, atherosclerosis, and coronary artery disease. N Eng J Med, 2005; 352:1685–1695.

Forsythe

, Wallace

JMW

, Livingstone

MBE

. Obesity and inflammation: The effects of weight loss. Nutr Res Rev, 2008; 21:117–133.

Klempel

, Varady

. Reliability of leptin, but not adiponectin, as a biomarker for diet-induced weight loss in humans. Nutr Res, 2011; 69:145–154.

Ahima

. Revisiting leptin's role in obesity and weight loss. J Clin Invest, 2008; 118:2380–2383.

Sacks

, Bray

, Carey

et al. Comparison of weight-loss diets with different compositions of fat, protein, and carbohydrates. N Engl J Med, 2009; 360:859–873.

Ebbeling

, Leidig

, Feldman

et al. Effects of a low-glycemic load vs low-fat diet in obese young adults: A randomized trial. Nat Clin Pract Endocrinol Metab, 2007; 3:802–803.

Volek

, Phinney

, Forsythe

et al. Carbohydrate restriction has a more favorable impact on the metabolic syndrome than a low fat diet. Lipid, 2009; 44:297–309.

10.

De Luis

, Aller

, Izaola

et al. Effects of a low-fat versus a low-carbohydrate diet on adipocytokines in obese adults. Horm Res, 2007; 67:296–300.

11.

Miller

, Volpe

, Coleman-Kelly

et al. Anthropometric and leptin changes in women following different dietary approaches to weight loss. Obesity, 2008; 17:199–201.

12.

Puglisi

, Fernandez

. Modulation of C-reactive protein, tumor necrosis factor-α, and adiponectin by diet, exercise, and weight loss. J Nutr, 2008; 138:2293–2296.

13.

Klemsdal

, Holme

, Nerland

et al. Effects of a low glycemic load diet versus a low-fat diet in subjects with and without the metabolic syndrome. Nutr Metab Cardiovasc Dis, 2010; 20:195–201.

14.

National Cholesterol Education Program (NCEP) Expert Panel on Detection Evaluation, Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation, 2002; 106:3143–3421.

15.

Seshadri

, Samaha

, Stern

et al. Adipocytokine changes caused by low-carbohydrate compared to conventional diets in obesity. Metab Syndr Rel Dis, 2005; 3:66–74.

16.

Lee

, Shin

, Choue

. Shifts in diet from high fat to high carbohydrate improved levels of adipokines and pro-inflammatory cytokines in mice fed a high-fat diet. Endocr J, 2010; 57:39–50.

17.

Madsen

, Rissanen

, Bruun

et al. Weight loss larger than 10% is needed for general improvement of levels of circulating adiponectin and markers of inflammation in obese subjects: A 3-year weight loss study. Eur J Endocrinol, 2008; 158:179–187.

18.

Lien

, Haqq

, Arlotto

et al. The STEDMAN Project: Biophysical, biochemical and metabolic effects of a behavioral weight loss intervention during weight loss, maintenance, and regain. OMICS, 2009; 13:21–35.

19.

Mather

, Funahashi

, Matsuzawa

et al. for the Diabetes Prevention Program. Adiponectin, change in adiponectin, and progression to diabetes in the Diabetes Prevention Program. Diabetes, 2008; 57:980–986.

20.

Reilly

, Lehrke

, Wolfe

et al. Resistin is an inflammatory marker of atherosclerosis in humans. Circulation, 2005; 111:932–939.

21.

Malo

, Ukkola

, Jokela

et al. Resistin is an indicator of the metabolic syndrome according to five different definitions in the Finnish health 2000 Survey. Metab Syndr Rel Dis, 2011; 9:203–210.

22.

Fargnoli

, Sun

, Olenczuk

et al. Resistin is associated with biomarkers of inflammation while total and high-molecular weight adiponectin are associated with biomarkers of inflammation, insulin resistance, and endothelial function. Eur J Endocrinol, 2010; 162:281–288.

23.

You

, Berman

, Ryan

et al. Effects of hypocaloric diet and exercise training on inflammation and adipocyte lipolysis in obese postmenopausal women. J Clin Endocrinol Metab, 2004; 89:1739–1746.

24.

Browning

, Krebs

, Magee

et al. Circulating markers of inflammation and their link to indices of adiposity. Obes Facts, 2008; 1:259–265.

25.

Sharman

, Volek

. Weight loss leads to reduction in inflammatory biomarkers after a very-low-carbohydrate diet and a low-fat diet in overweight men. Clin Sci, 2004; 107:365–369.

26.

Pears

, Rankin

. Inflammatory response to a high fat, low carbohydrate weight loss diet: effect of antioxidants. Obesity, 2008; 16:1573–1578.

27.

Ziccardi

, Nappo

, Giugliano

et al. Reduction of inflammatory cytokine concentrations and improvement of endothelial functions in obese women after weight loss over one year. Circulation, 2002; 105:804–809

28.

Clifton

, Keogh

, Foster

et al. Effect of weight loss on inflammatory and endothelial markers and FMD using two low-fat diets. Int J Obes, 2005; 29:1445–1451.

29.

Keogh

, Brinkworth

, Noakes

et al. Effects of weight loss from a very low-carbohydrate diet on endothelial function and markers of cardiovascular disease risk in subjects with abdominal obesity. Am J Clin Nutr, 2008; 87:567–576.

30.

Christiansen

, Richelsen

, Bruun

. Monocyte chemoattractant protein-1 is produced in isolated adipocytes, associated with adiposity and reduced after weight loss in morbid obese subjects. Int J Obes, 2005; 29:146–150.