Postprandial response of plasma IL-6 to isoenergetic meals rich in casein orpotato singly and combined in obese women
Abstract
Milk consumption decreases inflammatory stress in overweight and obese subjects. Caseinis the major protein in milk and enhances the secretion of insulin that hasanti-inflammatory activity. The aim of the present study was to compare the acute effectof meals rich in casein and carbohydrate and a combination of both nutrients onpostprandial plasma concentrations of IL-6, a marker of inflammation, in obese women. Atotal of twenty-five obese women aged 38–68 years consumed isoenergetic meals rich inpotato (POT) or casein (CA) or a combination of both these meals (POT + CA), in randomorder in a cross-over trial. After an overnight fast, blood samples were collected beforeand at 1 and 4 h after the meals and circulating concentrations of IL-6, glucose, insulinand NEFA were measured. Plasma IL-6 concentrations increased significantly(P < 0·001) during 4 h after the meals. The AUC of postprandialIL-6 concentrations was not significantly (P = 0·77) different among themeals. Postprandial serum insulin concentration AUC was significantly higher during thePOT + CA meal compared with the POT meal (P = 0·001) and the CA meal(P < 0·05), which in turn was significantly higher than the POTmeal (P < 0·05). These data show that while ingestion of CA aloneor combined with POT acutely increases circulating insulin concentrations, it does notappreciably alter the postprandial increase in plasma IL-6 concentrations in obesewomen.
IL-6 is a multifunctional cytokine that is synthesised in several tissues, and regulatesinnate immunity, the acute-phase response and central and peripheral nutrienthomeostasis(1). Under non-inflammatory conditions, adipose tissue supplies approximately30 % of circulating IL-6(2). In obese individuals, concentrations of IL-6 in fastingplasma(3,4) and adipose tissue(3) and release of IL-6 from adipose tissue into thecirculation(2) are abnormally high. These elevated levels of IL-6 are thought to reflectthe chronic, subclinical inflammation that is associated with obesity as a result of increasednumbers of macrophages in adipose tissue(5). Ingestion of food acutely increases IL-6 levels in adiposetissue(6) and plasma(7–10) and increases the release of IL-6 from skeletal muscle(11).
Milk products are an important source of protein in the Western diet. Consumption of low-fatdairy products is inversely associated with the risk of developing type 2diabetes(12). Casein accounts for 80 % of milk proteins and diets rich in casein seemto decrease body weight in obese women(13). An increase in milk intake for 28 d decreases fasting plasma IL-6concentrations in overweight and obese individuals(14). Ingestion of casein, like other proteins, enhances the secretion ofinsulin that is known to inhibit inflammation(15,16). Few if any studies have examined the acute effect of consuming meals richin casein and casein plus carbohydrate on postprandial plasma IL-6 concentrations in obesesubjects.
The aim of the present study was to compare the acute effects of isoenergetic mealscontaining casein or carbohydrate or in combination on plasma IL-6 concentrations in obesewomen.
Subjects and methods
Subjects
A total of twenty-five women with BMI ≥ 30 kg/m2 and aged 38–68 years,including eleven who did not have serious illnesses and were not receiving any medicationsand fourteen who were receiving prescribed medications for hypertension(n 8) and depression (n 8), were recruited. The presentstudy was conducted according to the guidelines laid down in the Declaration of Helsinkiand all procedures involving human subjects/patients were approved by the Lower SouthRegional Ethics Committee. Written and informed consent was obtained from allparticipants.
Study design
The study had a single-blind, randomised, cross-over design. Participants were randomlyassigned to a sequence of three test meals using the second generator on the
Meals
The potato (POT) meal contained 20 g dried potato flakes (Cinderella) that wasreconstituted into mashed potato by the addition of hot water (80 ml). The casein (CA)meal contained 19·5 g of sodium caseinate (Fonterra), 2·5 g cocoa powder and 1·5 teaspoonsof saccharin dissolved in 150 ml water. Consumption of both the POT meal and the CA mealat the same time constituted the POT + CA meal. The composition of meals is shown in Table 1.
| Meal… | POT | CA | POT + CA |
|---|---|---|---|
| Energy (kJ) | 309 | 308 | 617 |
| Carbohydrate (g) | 18·0 | 0·05 | 18·0 |
| Protein (g) | 1·3 | 18·0 | 19·0 |
| Fat (g) | 0·1 | 0·15 | 0·25 |
Laboratory methods
Venous blood was taken into tubes containing EDTA, fluoride and into plain tubes. Serumand EDTA plasma were separated by centrifugation of the tubes at 1500 g for 15 min at 4°C. Samples of serum and plasma were harvested and stored at –80°C.Plasma glucose was measured in fluoride anti-coagulated blood by routine automated methodsin the laboratories of Dunedin Public Hospital. Serum insulin was measured on a Hitachi911 autoanalyser using a commercial kit and calibrator (Roche Diagnostics). Plasma IL-6concentrations were measured in duplicate by sensitive enzyme-linked immunosorbent assayusing a commercial kit (R&D Systems). The intra-assay CV for this assay was 7 %.Samples from an individual were measured in the same assay to reduce inter-assayvariation.
Statistical analyses
Data are presented as mean values and standard deviations unless stated otherwise. Datawere log-transformed before statistical analysis using the IBM SPSS statistical software,version 20 (IBM Corp.). The trapezium method was used to calculate AUC(17). Repeated-measures ANOVA with simple within-subject contrasts was usedto compare AUC among the meals. Models were also tested with medication status as abetween-subjects factor. Repeated-measures ANOVA was also used to analyse changes invariables with time after meals and to estimate carry-over by comparing zero-time valuesamong the three visits. Two-sided tests of significance were used and a Pvalue of less than 0·05 was considered to be statistically significant.
Results
The characteristics of the obese women who participated in the study are shown in Table 2. Baseline concentrations of IL-6, glucose,insulin and NEFA in the circulation were higher compared with values (IL-6: 1·4 (sd0·6) ng/l; glucose: 4·6 (sd 0·40) mmol/l; insulin: 27 (sd 10) pmol/l;NEFA: 0·41 (sd 0·19) mmol/l) in fourteen lean women of comparable age (53(sd 10) years) who participated in a previous study from thislaboratory(9).
Baseline characteristics of the participants (n 25) determined atthe first visit
(Mean values and standard deviations)
| Characteristic | Mean | sd |
|---|---|---|
| Age (years) | 54 | 9 |
| BMI (kg/m2) | 35·9 | 5·0 |
| Waist (cm) | 109 | 10 |
| Glucose (mmol/l) | 5·5 | 0·4 |
| Insulin (pmol/l) | 77 | 60 |
| HOMA-IR | 1·45 | 1·11 |
| NEFA (mmol/l) | 0·45 | 0·17 |
| IL-6 (ng/l) | 1·93 | 1·25 |
Fig. 1 shows postprandial circulatingconcentrations and AUC of glucose, insulin, NEFA and IL-6 in the obese women during themeals. Serum insulin concentrations increased significantly(P < 0·001) during the meals. The AUC of postprandial serum insulinconcentrations was significantly (P < 0·05) higher during the CAmeal compared with the POT meal and was significantly higher during the POT + CA mealcompared with both the POT meal (P = 0·001) and the CA meal(P < 0·05). The AUC of postprandial NEFA concentrations wassignificantly (P < 0·01) higher during the CA meal compared with thePOT meal. The AUC of postprandial plasma glucose and IL-6 concentrations were notsignificantly different among the meals. There were no significant(P = 0·43–0·93) interactions between medication status and type of meal inAUC data. Zero-time circulating concentrations of glucose (P = 0·73),insulin (P = 0·77), NEFA (P = 0·14) and IL-6(P = 0·51) were not significantly different among the three visits.
Postprandial circulating concentrations of glucose (a), insulin (c), NEFA (e) andIL-6 (g) and AUC of glucose (b), insulin (d), NEFA (f) and IL-6 (h) during meals richin potato ( □ ; POT), casein (CAS; ◊) and potato + casein (POT + CA; ●) in obese women(n 25). Values are means, with standard errors represented byvertical bars. Concentrations of glucose, insulin, NEFA and IL-6 all changedsignificantly with time after meal (P < 0·001) inrepeated-measures ANOVA of log-transformed data. *Mean value was significantlydifferent from those of other meals (P < 0·05; within-subjectcontrasts in repeated-measures ANOVA of log-transformed data). Mean value wassignificantly different from that of the POT meal: †† P = 0·01, †††P = 0·001 (within-subject contrasts in repeated-measures ANOVA oflog-transformed data).
Discussion
Our data show that the postprandial increase in plasma IL-6 concentrations was similarafter ingestion of isoenergetic amounts of POT and CA and a combination of both nutrients inobese women. This finding is in keeping with the results of a previous study which reportedsimilar postprandial increases in plasma IL-6 concentrations after the consumption of mixedmeals rich in protein (that was derived from soya and whey), carbohydrate, or fat insubjects with the metabolic syndrome(10). Altogether, these studies suggest that the type of macronutrient andthe type of protein consumed does not differentially affect the postprandial increase inplasma IL-6 concentrations.
There is evidence that insulin has anti-inflammatory activity, including a decrease inmononuclear cell NF-κB during euglycaemic hyperinsulinaemia(15,16). On the other hand, an increase in plasma IL-6 concentrations atapproximately 4 h during a euglycaemic–hyperinsulinaemic clamp in healthymen(18), in subjects with type 2 diabetes and non-diabeticindividuals(19) has been reported. In the present study, postprandial serum insulinconcentrations were higher after ingestion of CA and even more so after ingestion ofPOT + CA compared with ingestion of POT alone while the response of plasma IL-6 levels didnot differ appreciably among these meals. It is possible that these postprandial increasesin serum insulin concentrations were not large enough to influence plasma IL-6concentrations. In previous studies, supraphysiological concentrations of circulatinginsulin were achieved during euglycaemic–hyperinsulinaemic clamps that increased plasma IL-6concentrations(18,19). A greater increase in postprandial serum insulin concentrations whencasein is added to a meal has been reported previously(20). Ingestion of milk and other food protein stimulate insulin secretionand increase insulin concentrations in the blood(21,22).
Postprandial hyperinsulinaemia inhibits adipose tissue lipolysis and NEFA release. In thepresent study, postprandial NEFA response, as indicated by AUC, was unexpectedly higherfollowing the ingestion of casein compared with potato despite a concomitantly largerincrease in serum insulin concentrations after intake of casein. It is possible that gastricemptying was more rapid after the CA meal compared with the other meals and the nadir ofpostprandial NEFA concentrations may have been earlier than 1 h and therefore undetected.The CA meal was liquid and would be emptied more rapidly from the stomach compared with theother meals that contained solid nutrients.
The meals in the present study had low energy content. Our preliminary studies in a smallnumber of obese women found that they were unable to comfortably consume a POT + CA mealwith twice the current amounts of these nutrients. Also, the amounts of protein andcarbohydrate were comparable with amounts used in previous studies(21,23). Investigation of the effect of meals low in energy content in obesesubjects is appropriate as they are advised to consume less food in order to lose weight.
The metabolic effect of the postprandial increase in plasma IL-6 is uncertain. There isevidence that IL-6 can affect glucose and lipid metabolism(1). Recently, it has been suggested that the postprandial increase inplasma IL-6 may be due, at least in part, to enhanced skeletal muscle expression of the IL-6gene and may be a normal, physiological response aimed at enhancing glucoseuptake(24).
The present study has a number of limitations. The number of subjects studied wasrelatively small. Thus, caution must be exercised in extrapolation of the findings to largerpopulations. The physical state of the meals was not identical. Thus, gastric emptying mayhave differed among the meals and influenced postprandial concentrations of measuredvariables. The proportions of carbohydrate and protein were reduced in the POT + CA mealcompared with the other meals and this may alter some metabolic responses. Some of the womenwere taking medications. However, medication use did not appear to affect postprandialresponses in our data. We did not study non-obese controls and cannot therefore directlyassess the effect of obesity on our findings. The number of postprandial measurements waslimited and this may also have limited the assessment of early changes in plasma insulin,glucose and NEFA after the meals. However, values of insulin and glucose at the currentpostprandial time points were comparable with those reported previously in healthy subjectsafter meals containing comparable amounts of protein and carbohydrate. After these meals,glucose concentrations were below baseline from 60 to 120 min(23). Postprandial hyperinsulinaemia can increase the disposal of bloodglucose so that it becomes greater than the absorption of glucose from the gut, leading to adecrease in blood glucose below baseline concentrations(25).
In conclusion, these data suggest that while ingestion of CA alone or combined with POTacutely increases postprandial insulin concentrations, it does not noticeably affect thepostprandial increase in plasma concentrations of IL-6.
Acknowledgements
The present study was supported by a grant from the Healthcare Otago Research Trust.
P. J. M. was responsible for the concept of the study, its design and writing themanuscript. W. H. F. S. was responsible for conducting the study, analysis of the data andwriting the manuscript. S. A. de J., A. R. R. and E. A. B. were responsible for conductingthe study. The authors are grateful to the participants in the study.
We acknowledge with sadness the recent death of our friend and colleague Sylvia De Jong.
The authors declare no conflict of interest.
References
- 1. & (2008) Role of interleukin-6 signalling in glucose andlipid metabolism. Acta Physiol (Oxf)192, 37–48[PubMed][Google Scholar]
- 2. , , ,
(1997) Subcutaneous adipose tissue releasesinterleukin-6, but not tumor necrosis factor-α, invivo. J Clin Endocrinol Metab82, 4196–4200[PubMed][Google Scholar] - 3. , , ,
(2006) Increased plasma interleukin-6concentrations and exaggerated adipose interleukin-6 content in severely obese patientsafter operative trauma. Surgery140, 50–67[PubMed][Google Scholar] - 4. , , ,
(2008) Visceral obesity and plasmaglucose-insulin homeostasis: contributions of interleukin-6 and tumor necrosis factor-αin men. J Clin Endocrinol Metab93, 1931–1938[PubMed][Google Scholar] - 5. , , ,
(2003) Obesity is associated with macrophageaccumulation in adipose tissue. J Clin Invest112, 1796–1808[PubMed][Google Scholar] - 6. , , ,
(1999) The differential effect of food intakeand β-adrenergic stimulation on adipose-derived hormones and cytokines inman. J Clin Endocrinol Metab84, 2126–2133[PubMed][Google Scholar] - 7. , , ,
(2002) Postprandial endothelial activation inhealthy subjects and in type 2 diabetic patients: role of fat and carbohydratemeals. J Am Coll Cardiol39, 1145–1150[PubMed][Google Scholar] - 8. , , ,
(2007) A high fat meal is accompanied byincreased plasma interleukin-6 concentrations. Nutr MetabCardiovasc Dis17, 195–202[PubMed][Google Scholar] - 9. , , ,
(2008) Postprandial cytokine concentrationsand meal composition in obese and lean women. Obesity16, 2046–2052[PubMed][Google Scholar] - 10. , , ,
(2010) Postprandial metabolic and hormonalresponses of obese dyslipidemic subjects with metabolic syndrome to test meals, rich incarbohydrate, fat, or protein. Atherosclerosis210, 307–313[PubMed][Google Scholar] - 11. , , ,
(2005) Postprandial interleukin-6 releasefrom skeletal muscle in men with impaired glucose tolerance can be reduced by weightloss. J Clin Endocrinol Metab90, 5819–5824[PubMed][Google Scholar] - 12. , , ,
(2005) Diary consumption and risk of type 2diabetes. A prospective study. Arch Int Med165, 997–1003[PubMed][Google Scholar] - 13. , , ,
(2007) Soy compared to casein mealreplacement shakes with energy-restricted diets for obese women: randomized controlledtrial. Metabolism56, 280–288[PubMed][Google Scholar] - 14. , , ,
(2010) Effects of dairy compared with soy onoxidative and inflammatory stress in overweight and obese subjects.Am J Clin Nutr91, 16–22[PubMed][Google Scholar] - 15. , , ,
(2001) Insulin inhibits intranuclear factorκB and stimulates IκB in mononuclear cells in obese subjects: evidence for ananti-inflammatory effect?J Clin Endocrinol Metab86, 3257–3266[PubMed][Google Scholar] - 16. , & (2002) The anti-inflammatory and potentiallyanti-atherogenic effect of insulin: a new paradigm. Diabetologia45, 924–930[PubMed][Google Scholar]
- 17. , , ,
(1990) Analysis of serial measurements inmedical research. BMJ300, 230–235[PubMed][Google Scholar] - 18. , , ,
(2004) Insulin stimulates interleukin-6 andtumor necrosis factor-α gene expression in human subcutaneous adiposetissue. Am J Physiol Endocrinol Metab286, E234–E238[PubMed][Google Scholar] - 19. , , ,
(2009) Acute hyperinsulinemia raises plasmainterleukin-6 in both nondiabetic and type 2 diabetes mellitus subjects, and this effectis inversely associated with body mass index. Metabolism58, 860–866[PubMed][Google Scholar] - 20. , , ,
(2010) Acute effects of casein onpostprandial lipemia and incretin responses in type 2 diabetic subjects.Nutr Metab Cardiovasc Dis20, 101–109[PubMed][Google Scholar] - 21. , , ,
(1986) The serum insulin and plasma glucoseresponses to milk and fruit products in type 2 (non-insulin dependent)diabetes. Diabetologia29, 784–791[PubMed][Google Scholar] - 22. , , ,
(1992) Metabolic response to cottage cheeseor egg white protein, with or without glucose, in type II diabeticsubjects. Metabolism41, 1137–1145[PubMed][Google Scholar] - 23. , , ,
(2004) Glycemia and insulinemia in healthysubjects after lactose-equivalent meals of milk and other food proteins: the role ofplasma amino acids and incretins. Am J Clin Nutr80, 1246–1253[PubMed][Google Scholar] - 24. , , ,
(2012) Inflammatory and oxidative stressresponses to high-carbohydrate and high-fat meals in healthy humans.J Nutr Metab2012, 238056.[Google Scholar] - 25. , & (1995) Physiological modulation of plasma free fattyacid concentrations by diet. Metabolic implications in nondiabeticsubjects. Diabetes Care18, 962–969[PubMed][Google Scholar]