BMJ Open. For peer review only -

Transcription

1 Ethnic differences in the association between cytokines and insulin sensitivity: A cross-sectional study comparing Iraqi immigrants with native Swedes Journal: BMJ Open Manuscript ID: bmjopen-0-00 Article Type: Research Date Submitted by the Author: -Jun-0 Complete List of Authors: Memon, Ashfaque; Lund University, Center for Primary Health Care Research Sundquist, Jan; Lund University, Center for Primary Health Care Research Wang, Xiao; Lund University, Center for Primary Health Care Research Palmer, Karolina; Lund University, Center for Primary Health Care Research Sundquist, Kristina; Lund University, Center for Primary Health Care Research Bennet, Louise; Lund University, Center for Primary Health Care Research <b>primary Subject Heading</b>: Secondary Subject Heading: Diagnostics Keywords: Diabetes and endocrinology DIABETES & ENDOCRINOLOGY, MOLECULAR BIOLOGY, General diabetes < DIABETES & ENDOCRINOLOGY BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

2 Page of BMJ Open Ethnic differences in the association between cytokines and insulin sensitivity: A crosssectional study comparing Iraqi immigrants with native Swedes Ashfaque.A. Memon *, Jan Sundquist, Xiao Wang, Karolina Palmér, Kristina Sundquist and Louise Bennet. Department of Clinical Sciences, Malmö, Lund University, Sweden and Center for Primary Health Care Research, Lund University/Region Skåne, Sweden *Corresponding Author: Ashfaque A. Memon Center for Primary Health Care Research Wallenberg Laboratory, th floor Inga Marie Nilsson s gata Skåne University Hospital S-0 0 Malmö Sweden ashfaque.memon@med.lu.se Short title: Insulin sensitivity and cytokines according to ethnicity Word count: 0 Keyword: Cytokines, Insulin sensitivity, type diabetes mellitus, ethnicity - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

3 Page of ABSTRACT: Objectives: To investigate whether the associations between cytokines and insulin sensitivity differ in Swedish residents born in Iraq compared to residents born in Sweden. Design: Cross sectional study. Settings: Iraqi and Swedish origin residents of Rosengård area of Malmö, aged to years were randomly selected from the census register. Participants/Methods: (Iraqi n=, Swedish n=) participants agreed to participate in the study. participants dropped out (Iraqi n=, Swedish n=). Participants who had already been diagnosed with type diabetes mellitus (TDM), those who could not participate in an oral glucose tolerance test (OGTT) (Iraqi n=, Swedish n=) and those who had cold/fever (n=) at the time of blood sampling were excluded. In total, serum samples from individuals of Swedish (n=) and Iraqi (n=) origin were included. Serum concentrations of a panel of cytokines, comprising interleukin (IL)-β, IL-, IL-, IL-, IL-, IL-, IL- (p0), IL-, Interferon (IFN)-γ and tumor necrosis factor (TNF)-α were analyzed by Luminex multiplex assay. Results: In the whole study population, levels of all tested cytokines were inversely associated with insulin sensitivity index (ISI), independent of age, sex, body mass index (BMI), sedentary lifestyle and family history of TDM (p= 0.0). Interestingly, stratification of the study population according to country of birth showed a significant inverse association between all tested cytokines and ISI in Iraqi-born population (p 0.0). The association was independent of age, sex, BMI, sedentary lifestyle and family history of TDM. In contrast except for IL- (p= 0.0) no other tested cytokine was found to be significantly associated with ISI in the Swedish-born population (p 0.0). Conclusions: Our results show a stronger association between cytokines and ISI in Iraqi compared to Swedish-born population, indicating a different role for cytokines in relation to - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

4 Page of BMJ Open insulin sensitivity in two ethic groups. Article summary: Article focus: The prevalence of type diabetes mellitus (TDM) is higher in immigrants from the Middle East compared to native Swedes. Cytokines play a significant role in insulin resistance however their role in relation to insulin resistance in different ethnic groups is not well defined. Key messages: Compared to Swedes we show a stronger association between cytokines and lower insulin sensitivity in immigrants from Middle East. The association was independent of clinical parameters found to be different between two groups. Together, our results indicate that the molecular mechanism of the pathogenesis of TDM may differ between Iraqis and Swedes. Strengths and limitations To our knowledge, this is the first study of its kind in which role of cytokines in relation to insulin sensitivity is compared between Swedes and immigrants from Middle East. The main limitation of our study is the small sample, which means that the results need to be confirmed in a larger population. However, it should be noted that it was conducted as a hypothesis-generating study. - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

5 Page of INTRODUCTION Type diabetes mellitus (TDM) is a chronic metabolic disorder caused by defects in insulin secretion, insulin action, or both. If ineffectively controlled, the resulting chronic hyperglycemia is associated with numerous disabling complications,[]. The prevalence of TDM is increasing, and it is estimated that by 0, % of the world s population will have TDM, impaired fasting glucose (IFG) or impaired glucose tolerance (IGT),[]. In Sweden, the prevalence of TDM among non-european immigrants is estimated to be - times higher than among native Swedes,[]. In addition, TDM develops an average of years earlier in immigrants from the Middle East than in native Swedes, and Middle Eastern immigrants diagnosed with TDM more frequently have a family history of TDM,[]. The main hypothesized mechanisms of insulin resistance in TDM are oxidative stress, endoplasmic reticulum stress, amyloid deposition in the pancreas, ectopic lipid deposition in the muscle, liver and pancreas, and lipotoxicity and glucotoxicity,[, ]. Although it is difficult to determine which mechanism is the most important in an individual tissue, model or patient with TDM, all these stresses can be caused by overnutrition. It is important to note, however, that each of these stresses could either induce an inflammatory response or be associated with inflammation,[, ]. The association between inflammation and insulin resistance and future development of TDM is well established,[, ]. The production of tumor necrosis factor (TNF)-α by cells in the adipose tissue of rodents provided early evidence of a link between tissue inflammation and the pathogenesis of insulin resistance and TDM,[]. In addition, interleukin (IL)-β contributes to the glucose-induced impairment of β-cell function and apoptosis,[]. Furthermore, it has been shown in both a mouse model of obesity and humans that inflammatory processes occurring in adipocytes contribute to the systemic low-grade inflammation leading to insulin resistance and ultimately TDM,[, ]. - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

6 Page of BMJ Open Obesity is a major risk factor for TDM and is highly prevalent amongst immigrants from the Middle East,[]. Obesity is associated with low-grade chronic inflammation, which activates the innate immune system. This activation leads to the release of pro-inflammatory cytokines such as TNF-α, IL-β and IL-, which disrupt insulin homeostasis by blocking major anabolic cascades downstream of insulin signaling,[]. Elevated levels of pro-inflammatory cytokines characterize the early pre-clinical stages of TDM,[-]. Stronger family history, earlier onset of TDM and sedentary lifestyle are other highly prevalent risk factors for TDM in immigrants from the Middle East,[, ]. Even though the association between high levels of cytokines and insulin resistance has been extensively studied, few studies have been performed to elucidate the roles of inflammatory markers in various ethnic groups. In one such study from the USA, circulating TNF-α levels were found to differ according to ethnicity, with Hispanics having the highest levels and African-Americans the lowest,[]. Understanding inflammatory networks in the population groups at highest risk of developing TDM is important in the prediction of disease onset and the evaluation of strategies for disease prevention. The aim of this study was to investigate the relationship between cytokines and insulin sensitivity in two ethnically different groups: residents of the Swedish city of Malmö who were born in Sweden and Iraq. To our knowledge, this is the first study of its kind in which profiles of as many as cytokines were analyzed and compared in two ethnic groups. - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

7 Page of MATERIAL AND METHODS Subjects Men and women residents in Rosengård area of Malmö, aged to years and born in Sweden (Swedes) or Iraq (Iraqis) were randomly selected from a census register. (Iraqi n=, Swedish n=) participants agreed to participate were included in the study between the st of February and st of March 0. participants dropped out (Iraqi n=, Swedish n=). Patients who had already been diagnosed with TDM, those who could not participate in an oral glucose tolerance test (OGTT) (Iraqi n=, Swedish n=) and those who had cold/fever (n=) at the time of blood sampling were excluded from the study. In total, individuals were included (Iraqi, n=; Swedes, n=). Normal glucose tolerance was defined as a fasting plasma glucose (0 h sample) level of <. mmol/l and a plasma glucose level of <. mmol/l h after a g OGTT [0] (Iraqi, %; Swedes, %). An abnormal glucose tolerance test was defined as either a fasting plasma glucose level of >. mmol/l and <.0 mmol/l and/or a h glucose level of. mmol/l,[0] (Iraqi, %; Swedes, %). Insulin sensitivity index (ISI) was calculated from the OGTT data as described previously,[]. The procedures to be employed were explained to the patients and each participant signed a written informed consent form, provided blood samples and filled in a questionnaire. Participants were asked not to eat after pm the day before sample collection. Clinical variable assessment Systolic and diastolic blood pressure, body mass index (BMI), waist circumference (abdominal obesity), sedentary lifestyle, cholesterol, triglycerides, high-density lipoprotein (HDL), low-density lipoprotein (LDL) and HbAc were measured as described previously []. C-reactive protein (CRP) was measured by a commercial kit from Roche diagnostics, according to the manufacturer s instructions. - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

8 Page of BMJ Open Sample preparation and quantification of cytokines Samples were collected into a. ml serum separating tube and were allowed to clot for h at room temperature. They were then centrifuged for min at C and serum was aliquoted and stored at -0 C for further use. The Bio-Plex Precision Pro Human Cytokine -Plex Panel (Bio-Rad Inc., Hercules, CA), a high-sensitivity Luminex-based magnetic bead assay, was used to quantify IL-β, IL-, IL-, IL-, IL-, IL-, IL- (p0), IL-, IFN-γ and TNF-α in serum according to the manufacturer s instructions. Briefly, samples were diluted : in the sample diluent provided with the kit and incubated with magnetic beads coupled to specific antibodies. Cytokines were detected with premixed detection antibody. Beads were read on a Bio-Plex Suspension Array System and the data were analyzed using Bio-Plex Manager software (version.0). Absolute concentrations were measured from a standard curve generated from eight serially diluted standards provided with the kit. Each sample was analyzed in duplicate. Values are presented in pg/ml. Each run included controls of known concentration for each cytokine and a blank. Detection limits for IL-β, IL-, IL-, IL-, IL-, IL-, IL- (p0), IL-, IFN-γ and TNF-α were 0., 0., 0., 0., 0., 0., 0., 0., 0. and 0. pg/ml respectively. Inter-assay % CV was between.-. (values were calculated from six different experiments). Statistical analysis Differences in sample characteristics between Swedish-born subjects and first-generation immigrants from Iraq were tested using Student s t-test for continuous variables, chi-square tests for dichotomous variables and the Wilcoxon rank-sum test for the variables with a nonnormal distribution (Table ). Distribution of all cytokines is presented with median and IQR (interquartile range) and tested with the Wilcoxon rank-sum test (Table ). Linear regression was used to examine the associations between cytokine levels and ISI. Natural log transformation of ISI was used to obtain an approximately normal distribution. - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

9 Page of IL-, IL- (p0), IFN-γ and TNF-α were transformed with a natural log function due to their highly skewed distribution. IL- was dichotomized into below detection limit (< 0. pg/ml) values and above detection limits ( 0.), due to nearly 0% of values being below detection limit. It was possible to include the other cytokines untransformed. First, univariate analysis was performed to assess the association of each cytokine with ISI in the whole sample. Thereafter, multivariate linear regression was performed with ISI as the dependent variable and with adjustment for BMI, age, sex, sedentary lifestyle and family history of TDM (Table ). All analyses were then stratified by country of birth to examine whether the associations differed between individuals born in Sweden and those born in Iraq (Table ). To test the robustness of the results from the regression models we repeated the analysis with dichotomization of all the cytokines. STATA version (Stata Corp LP) was used for all statistical analyses. The study was performed according to the declaration of Helsinki,[]. The ethical committee at Lund University approved the study (approval no. 00/) and written informed consent was given by all the participants in the study after full explanation of the purpose and nature of all procedures used. - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

10 Page of BMJ Open RESULTS Table shows the anthropometric and metabolic characteristics of the study participants born in Iraq and Sweden. Consistent with earlier reported data from the same population-based study,[], median ISI was higher in the Swedes compared to the Iraqis. Swedes were older and had lower BMI compared to Iraqis. Mean HDL levels were higher in Swedes compared to Iraqis, whereas triglyceride levels were higher in Iraqis compared to Swedes. The prevalence of family history of TDM was significantly higher in Iraqis compared to Swedes. Mean systolic and diastolic blood pressures were higher in Swedes than in Iraqis. There were no significant differences in sex, sedentary lifestyle, hyperglycemia, cholesterol, LDL, HbAc, C-peptide, waist circumference and levels of CRP between the two groups (Table ). Table shows a summary of the distribution of the cytokines. There were no significant differences in median values between study participants born in Iraq and Sweden. In the overall population, univariate analysis showed that except IL- (p0) (p= 0.0) the levels of all analyzed cytokines had significant (p 0.0) inverse associations with ISI (Table ). Interestingly, after adjusting for age, sex, BMI, sedentary lifestyle and family history of TDM, levels of all analyzed cytokines were inversely associated with ISI (p 0.0) (Table ). Stratification of the data according to ethnicity showed strong inverse associations between levels of IL-β, IL-, IL-, IL-, IL-, IL- (p0), IL- and IFN-γ (p 0.0) and ISI in Iraqis. Non-significant trend towards association for IL- (p = 0.0), TNF-α (p = 0.0) and ISI in Iraqis was also observed (Table ). However after adjustment for age, sex, BMI, sedentary lifestyle and family history of TDM, all tested cytokines were found to be associated with ISI in the Iraqi population (p 0.0) (Table ). Sub-analysis of IL- and TNF-α by separately entering age, sex, BMI, sedentary life style and family history of TDM in the model, showed that both sedentary life style and family history of TDM had an effect on association between ISI, IL- and TNF-α (data not shown). On the contrary, in the Swedes - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

11 Page of no cytokine, except IL- (p = 0.0) was significantly associated with ISI in univariate analysis and after adjusting for age, sex, BMI, sedentary life style and family history of TDM (Table ). Interestingly, not only did the p-values differ between Iraqi and Swedes, but also all β- values were higher in Iraqis than in Swedes, even after adjusting (e.g. β = -0. for TNF-α in Iraqis compared to -0. in Swedish, compare adjusted β values between Iraq and Sweden, Table ). We also repeated the analysis with all cytokines dichotomized with the th percentile as cut-off in the regression models and the results remained the same (data not shown). - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

12 Page of BMJ Open DISCUSSION For the first time, we investigated the expression of a panel of cytokines in Swedish-born and Iraqi-born populations and their associations with insulin sensitivity. We found that levels of all cytokines tested in this study were strongly associated with lower insulin sensitivity in Iraqis, compared to Swedes after controlling for age, sex, BMI, sedentary lifestyle and family history of type diabetes mellitus (TDM). Ethnic differences exist in the prevalence of TDM and its complications,[]; however, no previous study has investigated molecular differences in Middle Eastern immigrants and native Scandinavians. The prevalence of TDM in the immigrant population in Sweden has been shown to be higher compared to that in native Swedes,[] The majority of the immigrants in Sweden are from the Middle East and Iraqis constitute the largest immigrant population living in Malmö,[]. Consistent with our results, a previous study of patients born in the Middle East and Sweden suggests that patients born in the Middle East have earlier TDM onset, higher BMI and higher prevalence of family history of TDM compared to Sweden born patients with TDM,[]. Low-grade systemic inflammation results in insulin resistance, to which pancreatic islets respond by enhancing their cell mass and thereby their insulin secretory activity. However, exhaustion of this mechanism leads to insulin deficiency and ultimately TDM,[]. Elevated levels of cytokines such as IL- and TNF-α have been associated with TDM in other ethnic groups,[]. Iraqis are known to have lower insulin sensitivity compared to Swedes,[]. However, it is not known whether cytokines have any role in this difference. Our results show that high levels of all cytokines analyzed in this study (IL-β, IL-, IL-, IL-, IL-, IL-, IL- (p0), IL-, IFN-γ and TNF-α) were independently associated with lower insulin sensitivity in immigrants from Iraq, whereas high levels of only IL- was found to be associated with lower insulin sensitivity in Swedes. These results suggest that role of - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

13 Page of cytokines in insulin resistance may differ between Iraqis and Swedes. Obesity increases pro-inflammatory cytokine levels in both the circulation and peripheral tissues. Iraqi participants had higher BMI values than the Swedes,[]. However, the association between cytokines and insulin resistance in the Iraqi population was found to be independent of BMI, suggesting that the association between cytokines and insulin resistance was not due to the higher BMI values in the Iraqis. There was no difference in waist circumference between the two groups. It has been reported that, it is not the general adiposity, but rather visceral fat, which is associated with the incidence of prediabetes and TDM,[]. In fact, ectopic fat deposition in the abdominal viscera has been shown to be an important mediator of diabetes development than the total fat mass in obese individuals,[]. High visceral and hepatic fat levels have been associated with higher levels of IL-,[] and IL-,[0], respectively. Therefore the association between cytokine and insulin resistance in our study could be due to higher visceral or hepatic fat levels in the Iraqi population; however, we did not measure visceral or hepatic fat in our study population to prove this assumption. Family history of TDM was more prevalent in Iraqis compared to Swedes,[]. Interestingly, family history did not attenuate the association between cytokine levels and insulin resistance in our study, suggesting that the difference in the association between cytokines and insulin resistance between Iraqis and Swedes is independent of the high prevalence of family history of TDM in Iraqis. Together our results show that the association between cytokines and insulin resistance in Iraqis is not due to common clinical factors known to be associated with inflammatory responses in TDM. Further investigation is therefore warranted in order to understand the mechanism underlying the association between cytokines and insulin sensitivity in Iraqis. The main limitation of our study is the small sample, which means that the results need to be - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

14 Page of BMJ Open confirmed in a larger population. However, it should be noted that it was conducted as a hypothesis-generating study. In conclusion, we have shown for the first time the differences in cytokine levels between Iraqi- and Swedish-born residents of Sweden in relation to insulin sensitivity. We found a stronger inverse association between cytokines and insulin sensitivity in Iraqis compared to Swedes. This association was independent of sex, BMI, age, sedentary lifestyle or prevalence of family history of TDM. Our results suggest that the molecular mechanisms of the pathogenesis of TDM may differ between Iraqis and Swedes, and that Iraqi may need different intervention strategies to prevent and treat the disease. However, further investigation on a larger population is required to confirm these results. - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

15 Page of ACKNOWLEDGMENT We would like to thank Anna Hedelius for excellent technical support and Stephen Gilliver for critical reading of the article. Grant support The study was supported by ALF funding, FoU grants from Region Skåne and a grant from the Swedish Society of Medicine awarded to LB, and by grants awarded to JS and KS by the Swedish Research Council and an ALF project grant, Lund, Sweden. Author Contribution: AAM, JS, KS and LB conceived and designed the study; AAM and XW performed the experiments; AAM, JS, KP, KS and LB performed the data analysis and interpretation; and AAM, JS, XW, KP, KS and LB drafted and revised the article, and approved the final version. Authors declare no conflict of interest - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

16 Page of BMJ Open REFERENCES. Clark CM, Jr., Lee DA. Prevention and treatment of the complications of diabetes mellitus. The New England journal of medicine ; :-.. Zimmet P, Magliano D, Matsuzawa Y, et al. The metabolic syndrome: a global public health problem and a new definition. J Atheroscler Thromb 00; :-00.. Wandell PE, Johansson SE, Gafvels C, et al. Estimation of diabetes prevalence among immigrants from the Middle East in Sweden by using three different data sources. Diabetes & metabolism 00; :-.. Glans F, Elgzyri T, Shaat N, et al. Immigrants from the Middle-East have a different form of Type diabetes compared with Swedish patients. Diabet Med 00; :0-0.. Harding HP, Ron D. Endoplasmic reticulum stress and the development of diabetes: a review. Diabetes 00; Suppl :S-.. Weir GC, Bonner-Weir S. Five stages of evolving beta-cell dysfunction during progression to diabetes. Diabetes 00; Suppl :S-.. Donath MY, Schumann DM, Faulenbach M, et al. Islet inflammation in type diabetes: from metabolic stress to therapy. Diabetes care 00; Suppl :S-.. Hotamisligil GS, Erbay E. Nutrient sensing and inflammation in metabolic diseases. Nature reviews Immunology 00; :-.. Donath MY, Boni-Schnetzler M, Ellingsgaard H, et al. Islet inflammation impairs the pancreatic beta-cell in type diabetes. Physiology (Bethesda) 00; :-.. Shoelson SE, Lee J, Goldfine AB. Inflammation and insulin resistance. The Journal of clinical investigation 00; :-0.. Hotamisligil GS, Shargill NS, Spiegelman BM. Adipose expression of tumor necrosis factor-alpha: direct role in obesity-linked insulin resistance. Science ; : BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

17 Page of Maedler K, Sergeev P, Ris F, et al. Glucose-induced beta cell production of IL-beta contributes to glucotoxicity in human pancreatic islets. The Journal of clinical investigation 00; :-0.. Bastard JP, Maachi M, Lagathu C, et al. Recent advances in the relationship between obesity, inflammation, and insulin resistance. European cytokine network 00; :-.. Schmidt MI, Duncan BB, Sharrett AR, et al. Markers of inflammation and prediction of diabetes mellitus in adults (Atherosclerosis Risk in Communities study): a cohort study. Lancet ; :-.. Bennet L, Johansson SE, Agardh CD, et al. High prevalence of type diabetes in Iraqi and Swedish residents in a deprived Swedish neighbourhood--a population based study. BMC public health 0; :0.. Festa A, D'Agostino R, Jr., Tracy RP, et al. Elevated levels of acute-phase proteins and plasminogen activator inhibitor- predict the development of type diabetes: the insulin resistance atherosclerosis study. Diabetes 00; :-.. Meigs JB, Hu FB, Rifai N, et al. Biomarkers of endothelial dysfunction and risk of type diabetes mellitus. JAMA 00; :-.. Pradhan AD, Manson JE, Rifai N, et al. C-reactive protein, interleukin, and risk of developing type diabetes mellitus. JAMA : the journal of the American Medical Association 00; :-.. Olson NC, Callas PW, Hanley AJ, et al. Circulating levels of TNF-alpha are associated with impaired glucose tolerance, increased insulin resistance, and ethnicity: the Insulin Resistance Atherosclerosis Study. The Journal of clinical endocrinology and metabolism 0; : BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

18 Page of BMJ Open Organization WH. Defination, diagnosis and classification of diabetes mellitus and its complications: Report of a WHO consultation. Part. Diagnosis and classification of of diabetes mellitus. Geneva: World Health Organization.. Matsuda M, DeFronzo RA. Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes care ; :-0.. WMA. Declaration of Helsinki - Ethical principals of medical research involving human subjects Abate N, Chandalia M. The impact of ethnicity on type diabetes. Journal of diabetes and its complications 00; :-.. Statistics Sweden.: Statistics Sweden.. Donath MY, Shoelson SE. Type diabetes as an inflammatory disease. Nature reviews Immunology 0; :-.. Mirza S, Hossain M, Mathews C, et al. Type -diabetes is associated with elevated levels of TNF-alpha, IL- and adiponectin and low levels of leptin in a population of Mexican Americans: a cross-sectional study. Cytokine 0; :-.. Neeland IJ, Turer AT, Ayers CR, et al. Dysfunctional adiposity and the risk of prediabetes and type diabetes in obese adults. JAMA : the journal of the American Medical Association 0; 0:0-.. McLaughlin T, Lamendola C, Liu A, et al. Preferential fat deposition in subcutaneous versus visceral depots is associated with insulin sensitivity. The Journal of clinical endocrinology and metabolism 0; :E-0.. van den Borst B, Gosker HR, Koster A, et al. The influence of abdominal visceral fat on inflammatory pathways and mortality risk in obstructive lung disease. The American journal of clinical nutrition 0; : BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

19 Page of Kim JS, Le KA, Mahurkar S, et al. Influence of elevated liver fat on circulating adipocytokines and insulin resistance in obese Hispanic adolescents. Pediatric obesity 0; : BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

20 Page of BMJ Open Table. Baseline anthropometric and metabolic characteristics of the study population by country of birth Covariate Iraq (n = ) Sweden (n = ) p-value ISI a () () 0.0 Sex (female/male) b / / 0. Age (years) c (.) (.) 0.0 BMI (kg/m ) c (.) (.) 0.00 Sedentary lifestyle (yes/no) b / / 0. Hyperglycemia (yes/no) b,d / / 0. Family history (yes/no) b / / Cholesterol (mmol/l) c. (0.). (.) 0.0 LDL (mmol/l) c. (0.). (.0) 0. HDL (mmol/l) c. (0.). (0.) 0.00 Triglycerides (mmol/l) a. (.0).0 (.) 0.0 HbAc mmol/mol (%) e 0 (.) 0 (.) 0. C-peptide (mmol/l) a 0. (0.) 0. (0.) 0. Systolic BP (mmhg) c () () Diastolic BP (mmhg) c () () 0.00 Waist circumference (cm) c () () 0. C-reactive protein (mg/l) c. (.). (.) 0. a Insuline sensitivity index (ISI), triglycerides and C-peptide are presented as the median and interquartile range. b Sex, sedentary lifestyle, hyperglycemia and family history are presented as percentages. c Age, BMI (body mass index), cholesterol, LDL (low-density lipoprotein), HDL (high-density lipoprotein), systolic and diastolic BP (blood pressure) waist circumference and C-reactive protein are presented as the mean and standard deviation. d Hyperglycemia is defined as patients with impaired fasting glucose, impaired glucose tolerance or diabetes. e HbAc is presented as mean in IFCC units (NGSP units) - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

21 Page 0 of Table. Distribution of cytokine levels, stratified by country of birth. *Values are shown in pg/ml a P-value for test between Iraq and Sweden All (n = ) Iraq (n = ) Sweden (n = ) Cytokines* Median (IQR) Median (IQR) Median (IQR) p-value a IL-β 0. (0.) 0. (0.) 0. (0.) 0. IL-. (.). (.). (.) 0. IL- 0. (0.) 0. (0.) 0. (0.) 0. IL-. (.). (.). (.) 0. IL-. (.). (.) 0 (.) 0. IL-. (.). (.). (.) 0. IL-(p0). (.). (.).0 (.) 0. IL-.0 (.).0 (.).0 (.) 0. IFN-γ. (.). (.). (.) 0. TNF-α 0. (0.) 0. (0.) 0. (0.) 0. on January 0 by guest. Protected by copyright. - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from 0

22 Page of BMJ Open Table. Multivariate linear regression of the whole study population with logarithmic insulin sensitivity index as the dependent variable and cytokine level as the exposure Cytokines a Unadjusted Adjusted b β p % CI β p % CI IL-β -0. < , , -0.0 IL < , < , -0.0 IL < , , -0.0 IL , , -0.0 IL , , -0.0 IL , < , -0.0 IL- (p0) , , -0.0 IL- -0. < , , -0.0 IFN-γ , , 0.00 TNF-α , , -0.0 a IL-β, IL-, IL-, IL- and IL- were used untransformed, IL-, IL- (p0), IFN-γ and TNF-α were ln-transformed and IL- was dichotomized into below (< 0. pg/ml) and above ( 0.) detection limits. b Covariates in the models: age, sex, BMI, sedentary lifestyle and family history. - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

23 Page of Table. Multivariate linear regression by country of birth with logarithmic insulin sensitivity index as the dependent variable and cytokine level as the exposure Cytokines a Iraq Sweden Unadjusted Adjusted b Unadjusted Adjusted b β p % CI β p % CI β p % CI β p % CI IL-β -0.0 < , < , , , 0. IL < , < , , , 0.0 IL , < , , , 0. IL , < , , , 0. IL , , , , 0.00 IL , < , , , 0.0 IL- (p0) , , , , 0. IL- -0. < , < , , , 0.0 IFN-γ , , , , 0. TNF-α , , , , 0. a IL-β, IL-, IL-, IL- and IL- were used untransformed, IL-, IL- (p0), IFN-γ and TNF-α were ln-transformed and IL- was dichotomized into below (< 0. pg/ml) and above ( 0.) detection limits. b Covariates in the models: age, sex, BMI, sedentary lifestyle and family history. on January 0 by guest. Protected by copyright. - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from

24 The association between cytokines and insulin sensitivity in Iraqi immigrants and native Swedes Journal: BMJ Open Manuscript ID: bmjopen-0-00.r Article Type: Research Date Submitted by the Author: -Aug-0 Complete List of Authors: Memon, Ashfaque; Lund University, Center for Primary Health Care Research Sundquist, Jan; Lund University, Center for Primary Health Care Research Wang, Xiao; Lund University, Center for Primary Health Care Research Palmer, Karolina; Lund University, Center for Primary Health Care Research Sundquist, Kristina; Lund University, Center for Primary Health Care Research Bennet, Louise; Lund University, Center for Primary Health Care Research <b>primary Subject Heading</b>: Secondary Subject Heading: Diagnostics Keywords: Diabetes and endocrinology DIABETES & ENDOCRINOLOGY, MOLECULAR BIOLOGY, General diabetes < DIABETES & ENDOCRINOLOGY BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

25 Page of BMJ Open The association between cytokines and insulin sensitivity in Iraqi immigrants and native Swedes Ashfaque.A. Memon *, Jan Sundquist, Xiao Wang, Karolina Palmér, Kristina Sundquist and Louise Bennet. Center for Primary Health Care Research, Department of Clinical Sciences, Malmö, Lund University, Sweden and Region Skåne, Sweden *Corresponding Author: Ashfaque A. Memon Center for Primary Health Care Research Wallenberg Laboratory, th floor Inga Marie Nilsson s gata Skåne University Hospital S-0 0 Malmö Sweden ashfaque.memon@med.lu.se Short title: Insulin sensitivity and cytokines according to ethnicity Word count: Keyword: Cytokines, Insulin sensitivity, type diabetes mellitus, ethnicity - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

26 Page of ABSTRACT: Objectives: To investigate the associations between cytokines and insulin sensitivity in Swedish residents born in Iraq and Swedish residents born in Sweden. Design: Cross sectional study. Settings: Iraqi and Swedish origin residents of Rosengård area of Malmö, aged to years were randomly selected from the census register. Participants/Methods: (Iraqi n=, Swedish n=) participants agreed to participate in the study. participants dropped out (Iraqi n=, Swedish n=). Participants who had already been diagnosed with type diabetes mellitus (TDM), those who could not participate in an oral glucose tolerance test (OGTT) and those who had a cold/fever at the time of blood sampling were excluded. In total, serum samples from individuals of Swedish (n=) and Iraqi (n=) origin were included. Serum concentrations of a panel of cytokines, comprising interleukin (IL)-β, IL-, IL-, IL-, IL-, IL-, IL- (p0), IL-, Interferon (IFN)-γ and tumor necrosis factor (TNF)-α were analyzed by Luminex multiplex assay. Results: In the whole study population, levels of all tested cytokines were inversely associated with insulin sensitivity index (ISI), independent of age, sex, body mass index (BMI), sedentary lifestyle and family history of TDM (p= 0.0). Interestingly, stratification of the study population according to country of birth showed a significant inverse association between all tested cytokines and ISI in the Iraqi-born population (p 0.0). The association was independent of age, sex, BMI, sedentary lifestyle and family history of TDM. In contrast, with the exception for IL- (p= 0.0), no other tested cytokine was found to be significantly associated with ISI in the Swedish-born population (p 0.0). Conclusions: Our results show an association between cytokines and ISI and this association may be stronger in the Iraqi compared to the Swedish-born population. - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

27 Page of BMJ Open Article summary: Article focus: The prevalence of type diabetes mellitus (TDM) is higher in immigrants from the Middle East compared to native Swedes. Cytokines play a significant role in insulin resistance; however, their role in relation to insulin resistance in different ethnic groups is not well defined. Key messages: There is an association between cytokines and lower insulin sensitivity in immigrants from the Middle East but not in Swedes. The association is independent of clinical parameters. The molecular mechanisms in the pathogenesis of TDM may differ between Iraqis and Swedes. Strengths and limitations This is the first study in which the role of cytokines in relation to insulin sensitivity is examined in immigrants from the Middle East as well as in the native population of Swedes. The main limitation of our study is the relatively small sample size, which means any ethnic differences need to be confirmed in a larger population. However, it should be noted that the study was conducted as a hypothesisgenerating study. - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

28 Page of INTRODUCTION Type diabetes mellitus (TDM) is a chronic metabolic disorder caused by defects in insulin secretion, insulin action, or both. If ineffectively controlled, the resulting chronic hyperglycemia is associated with numerous disabling complications, []. The prevalence of TDM and related conditions is increasing worldwide, and it is estimated that by 0, % of the world s population will have TDM, impaired fasting glucose (IFG) or impaired glucose tolerance (IGT), [] In Sweden, the prevalence of TDM among non-european immigrants is estimated to be - times higher than among native Swedes, []. In addition, TDM develops an average of years earlier in immigrants from the Middle East than in native Swedes, and Middle Eastern immigrants diagnosed with TDM more frequently have a family history of TDM, []. The main hypothesized mechanisms of insulin resistance in TDM are oxidative stress, endoplasmic reticulum stress, amyloid deposition in the pancreas, ectopic lipid deposition in the muscle, liver and pancreas, and lipotoxicity and glucotoxicity, [, ]. Although it is difficult to determine which mechanism is the most important in the different tissues, model or patient with TDM, all these stresses can be caused by overnutrition. It is important to note, however, that each of these stresses could either induce an inflammatory response or be associated with inflammation, [, ]. The association between inflammation and insulin resistance and future development of TDM has been shown in previous research, [, ]. The production of tumor necrosis factor (TNF)-α by cells in the adipose tissue of rodents provided early evidence of a link between tissue inflammation and the pathogenesis of insulin resistance and TDM, []. In addition, interleukin (IL)-β contributes to the glucose-induced impairment of β-cell function and apoptosis, []. Family history and a sedentary lifestyle are highly prevalent risk factors for TDM in - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

29 Page of BMJ Open immigrants from the Middle East, [, ]. Obesity is another major risk factor for TDM and it is highly prevalent amongst immigrants from the Middle East, []. Obesity is associated with low-grade chronic inflammation, which activates the innate immune system. This activation leads to the release of pro-inflammatory cytokines such as TNF-α, IL-β and IL-, which disrupt insulin homeostasis by blocking major anabolic cascades downstream of insulin signaling, []. Elevated levels of pro-inflammatory cytokines may thus characterize the early pre-clinical stages of TDM, [-]. Even though the association between high levels of cytokines and insulin resistance has been well studied, few studies have been performed to elucidate the roles of inflammatory markers in various ethnic groups. In one such study from the USA, circulating TNF-α levels were found to differ according to ethnicity, with Hispanics having the highest levels and African- Americans the lowest, []. Characterizing inflammatory markers in the population groups at highest risk of developing TDM is important in the prediction of disease onset and the evaluation of strategies for disease prevention. The aim of this study was to investigate the relationship between cytokines and insulin sensitivity in two ethnically different groups: residents of the Swedish city of Malmö who were born in Sweden and Iraq. To our knowledge, this is the first study of its kind in which profiles of as many as cytokines were analyzed in two ethnic groups. - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

30 Page of MATERIAL AND METHODS Subjects Male and female residents in the Rosengård residential area of Malmö, aged to years and born in Sweden (Swedes) or Iraq (Iraqis) were randomly selected from the census register. We chose the ages to years since that was the age group of the non-retired population with the highest probability of identifying individuals with prediabetes or diabetes, since the prevalence of diabetes increases with age. In total, were invited to participate. Of these, participants (Iraqi n=, Swedish n=) agreed to participate and were included in the study between the st of February and st of March 0. Nineteen participants dropped out (Iraqi n=, Swedish n=). Patients who had already been diagnosed with TDM, those who could not participate in an oral glucose tolerance test (OGTT) (Iraqi n=, Swedish n=) and those who had a cold/fever (n=) at the time of blood sampling were excluded from the study. In total, individuals were included (Iraqi, n=; Swedes, n=). Normal glucose tolerance (NGT) was defined as a fasting plasma glucose (0 h sample) level of <. mmol/l and a plasma glucose level of <. mmol/l h after having been given g glucose in the OGTT, []. Impaired fasting glucose (IFG) was defined as a fasting plasma glucose level of >. mmol/l and <.0 mmol/l and a h glucose level of <. mmol/l. An impaired glucose tolerance (IGT) test was defined as either a fasting plasma glucose level of >. mmol/l and <.0 mmol/l and/or a h glucose level of. mmol/l and <. mmol/l. TDM was defined as fasting plasma glucose level of.0 mmol/l and/or OGTT of. mmol/l[]. Hyperglycemia was defined as patients with IFG, IGT or TDM. Insulin sensitivity index (ISI) was calculated from the OGTT data as described previously, []. The procedures to be employed were explained to the patients and each participant signed a written informed consent form, provided blood samples and filled in a questionnaire. - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

31 Page of BMJ Open Participants were asked not to eat after pm the day before the sample collection. The study was performed according to the declaration of Helsinki, [0]. The ethical committee at Lund University approved the study (approval no. 00/) and written informed consent was given by all the participants in the study after full explanation of the purpose and nature of all procedures used. Clinical variable assessment Leisure time physical activity (LTPA) was classified into four categories as described previously, [] Individuals in categories to were considered physically active while those in category were considered sedentary. Serum insulin levels were determined using the Radioimmunoassay technique (Access Ultrasensitive Insulin, Beckman Coulter, USA), []. A standard of g OGTT was performed and blood glucose was measured in blood samples collected at 0, 0, 0 and minutes by a HemoCue photometer (HemoCue AB, Ängelholm, Sweden), []. Cholesterol and triglycerides were analyzed using enzymatic methods whereas HDLcholesterol was measured after isolation of LDL and VLDL (Boehringer Mannheim GmbH, Germany) and LDL-cholesterol was analyzed using Friedewald s method, []. Systolic and diastolic blood pressures were measured twice in the supine position after minutes rest. The mean of the two measurements was calculated. Body mass index (BMI) was calculated as weight (kg) divided by height (m) squared. Abdominal obesity was defined as waist circumference of cm for males and cm for females, []. C-reactive protein (CRP) was measured by a commercial kit from Roche diagnostics, according to the manufacturer s instructions. Sample preparation and quantification of cytokines Samples were collected into a. ml serum separating tube and were allowed to clot for h at room temperature. They were then centrifuged for min at C and serum was aliquoted and - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.

32 Page of stored at -0 C for further use. The Bio-Plex Precision Pro Human Cytokine -Plex Panel (Bio-Rad Inc., Hercules, CA), a high-sensitivity Luminex-based magnetic bead assay, was used to quantify IL-β, IL-, IL-, IL-, IL-, IL-, IL- (p0), IL-, IFN-γ and TNF-α in serum according to the manufacturer s instructions. Briefly, samples were diluted : in the sample diluent provided with the kit and incubated with magnetic beads coupled to specific antibodies. Cytokines were detected with premixed detection antibody. Beads were read on a Bio-Plex Suspension Array System and the data were analyzed using Bio-Plex Manager software (version.0). Absolute concentrations were measured from a standard curve generated from eight serially diluted standards provided with the kit. Each sample was analyzed in duplicate. Values are presented in pg/ml. Each run included controls of known concentration for each cytokine and a blank. Detection limits for IL-β, IL-, IL-, IL-, IL-, IL-, IL- (p0), IL-, IFN-γ and TNF-α were 0., 0., 0., 0., 0., 0., 0., 0., 0. and 0. pg/ml, respectively. Inter-assay % CV was between.-. (values were calculated from six different experiments). Statistical analysis Differences in sample characteristics between Swedish-born subjects and first-generation immigrants from Iraq were tested using Student s t-test for continuous variables, chi-square tests for dichotomous variables and the Wilcoxon rank-sum test for the variables with a nonnormal distribution (Table ). Distribution of all cytokines is presented with median and IQR (interquartile range) and tested with the Wilcoxon rank-sum test (Table ). Table and were also adjusted for age, sex and BMI and adjusted p-values were calculated using linear regression for continuous variables, logistic regression for dichotomous variables, ordered logistic regression for normal/impaired/td and rank ANCOVA, [] for variables with nonnormal distribution. Linear regression was used to examine the associations between cytokine levels and ISI. - BMJ Open: first published as./bmjopen-0-00 on November 0. Downloaded from on January 0 by guest. Protected by copyright.