The association between circulating levels of myeloperoxidase and type 2 diabetes in the Malmö diet and cancer cohort. An Assessment

Table of contents

Abstract

List of abbreviations

Introduction

Public health relevance

Aim of the study

Objectives

Method and Material

Study sample

Consent

Identification of incident diabetes

Biochemical MPO analysis

Measurements and definition of variables

Statistical analysis

Results

Discussion

Strengths and limitations of the study

Clinical implications

Conclusion

References

List of figures and tables

Appendices

Popular science summary

Acknowledgement

List of abbreviations

Abbildung in dieser Leseprobe nicht enthalten

Abstract

Background

Chronic inflammation is suggested to play an important role in the pathogenesis of many chronic conditions. Chronic inflammatory biomarker Myeloperoxidase (MPO) has been reported in several studies that have investigated its role in cardiopulmonary events. However, there are limited studies that investigated its relationship with type 2 diabetes.

Objectives

To investigate the relationship between plasma MPO concentration and lifestyle and

socioeconomic factors and incident type 2 diabetes mellitus in the Malmö diet and cancer cohort (MDC).

Method

This was a prospective cohort study of 4544 participants drawn from MDC cardiovascular sub-cohort, with no type 2 diabetes at baseline between 1991 and 1996. Plasma MPO concentration were measured using Olink CVD kit. The cox proportional hazard regression analysis was used to determine the relationship of plasma MPO concentration with incident type 2 diabetes.

Results

Age, obesity, education level, Hb1Ac, plasma insulin, and plasma glucose were associated with MPO concentration. There was significant association between plasma MPO concentration with the risk of incident type 2 diabetes, crude hazard ratio (HR) 1.67 (95% CI

1.36-2.05) for higher plasma MPO concentration (quartile 4) compared with the lowest (quartile 1). This association persisted after further adjustment for age, sex, smoking, education level and BMI with a slight reduction in the risk, HR 1.48 (95% CI 1.2 - 1.82) for MPO quartile 4.

Conclusion

Our results suggest that age, obesity, education level, Hb1Ac, plasma insulin, and plasma glucose, are predictors of the chronic inflammatory biomarker plasma MPO concentration, which in turn are associated with risk of type 2 diabetes mellitus among females and males who participated in the MDC cardiovascular cohort. Before MPO can be used as a measure for preclinical type 2 diabetes or insulin resistance, its pathophysiological role needs to be further studied.

INTRODUCTION

According to the World Health Organisation (WHO), the estimated global diabetes prevalence in 2014 was 9% among adults 18 years or older and in 2012 diabetic attributable deaths were estimated at 1.5 million. More than 80% of these deaths occur in low and middle income countries and it was projected that it will be the 7th leading lone cause of death in 2030(1). Globally type 2 diabetes comprises 90% of the total diabetic patient population; type 2 diabetes is attributed to obesity and sedentary lifestyle(1). Unabated diabetes is associated with poor prognosis, resulting from morbidity and mortality which compromises the quality of life. It requires high costs to control diabetes and treatment of its acute and chronic complications(2). All diabetes prevalence in Sweden was estimated to be 6.8% in 2013 and projected to be 10.4% by the year 2050 affecting over 940000 individuals(3).

Diabetes is a chronic metabolic disorder that occurs when the pancreas does not produce insulin or when the body cannot effectively utilize the insulin produced(4). Insulin is a hormone required for the glucose in blood to be transformed into energy by the body cells. Type 2 diabetes is characterized by insulin resistance by target organs and less insulin production by the pancreas.

Inflammation is a protective response which involves host cells, blood vessels and proteins with primary intent to remove cause of injury, eradicate dead body cells and tissue and again start off the process of repair. It is a localized protective response to trauma or microbial invasion that destroys, dilutes or shields off the injurious agent and the injured tissue(5). The process is normally controlled and self-limited. It is characterised by redness, hotness, swelling, loss of function and pain often localized to the affected area. This process can occur in response to mechanical trauma, toxins and neoplasia. Excessive inflammation may cause abnormal recognition of host tissue as foreign or chronic inflammatory process may lead to inflammatory diseases such as type 2 diabetes, atherosclerosis and Alzheimer’s disease and syndromes such as rheumatic fever and rheumatic heart disease(5). The polymorphonuclear neutrophils (PMNs) and macrophages constitute part of the defensive white blood cells and are wide spread throughout the human body. They have a major role in innate immunity, their programmed cell death and removal is imperative in effective resolution of acute inflammation(6). However, these cells have azurophilic granules which contain an enzyme Myeloperoxidase (MPO) which is a haem protein released into extracellular fluid during inflammatory processes and is said to be associated with bacterial destruction and oxidative tissue injury(6,7). During inflammation there is activation of PMNs, a process referred to as respiratory burst, which leads to production and release of superoxide, hydrogen peroxide and other reactive oxygen species (ROS) which are toxic to microbes(8). The enzyme MPO catalyses the reaction between hydrogen peroxide and free chloride ions forming hypochlorous acid which is over fifty times more toxic than hydrogen peroxide to bacteria(8). It is believed that MPO plays a central role in killing fungi, protozoa, viruses, tumour cells, natural killer (NK) cells, red blood cells and blood platelets(8). Overproduction of ROS causes oxidative stress, hyperglycaemic states and creates an imbalance in glucose metabolism (9,10). Several studies have supported the hypothesis that chronic subclinical inflammation may be associated with insulin resistance and precede the development of clinically overt type 2 diabetes (11).

In a matched cohort study, it was found out that chronic organ specific and multisystem inflammatory conditions were associated with high risk of incidence type 2 diabetes(12). Recent findings show that concentrations of acute-phase response markers and mediators of inflammation cytokines such as Tumour Necrosis Factor α (TNFα) and interleukin-6 (ITN-6) are raised in people with type 2 diabetes(13). In a prospective multicentre case- control study, obesity was found to be characterized with chronic subclinical inflammation(14) and elevated concentration of MPO was observed in adipose tissue of animal models and humans. This enzyme plays an important role in the initiation and progression of chronic conditions such a Coronary Artery Disease (CAD)(14). It has been observed that MPO in blood is higher in patients with acute coronary syndrome (ACS) and it is not specific only for cardiac diseases, since activation of neutrophils and macrophages can be caused by several other conditions including infections (7). Recent findings suggest that elevated MPO blood levels among patients with ACS have poor prognostic outcomes(15). It has been highlighted that rising MPO levels among diabetic patients is greatly associated with progression of atherosclerosis(16). However, studies examining the association between MPO levels and risk of type 2 diabetes are lacking. Similarly, MPO could be both a direct and indirect early biomarker of systemic inflammation(17).

Public Health Relevance

Inflammation is a normal physiological process which enhances recovery and tissue repair after injury. If left unabated it can predispose one to getting future type 2 diabetes mellitus and its associated costs.

Aim of the study

To assess the association between chronic inflammation and incident type 2 diabetes in the Malmö diet and cancer cohort (MDC) by analysing chronic inflammatory MPO plasma biomarker. The MDC cohort is a large prospective population study where a wide array of clinical and non-clinical investigations are conducted to answer different research questions from a Swedish perspective.

Objectives

1. To investigate the relationship between MPO concentration and lifestyle and socioeconomic, i.e., obesity, smoking, physical activity and education.

2. To investigate the relationship between plasma MPO concentration with incident type 2 diabetes mellitus.

METHOD AND MATERIALS Study sample.

The Malmö Diet and Cancer Study (MDC) is a population-based, prospective cohort of 28449 men (born 1923-1945) and women (born 1923-1950) from Malmö, south of Sweden, who underwent baseline examinations between 1991 to 1996. At baseline, a detailed clinical examination including; dietary assessment, a self-administered questionnaire, anthropometry was done and peripheral venous blood samples drawn were archived in biological bank(18). From the main cohort, a subsample of 6,103 participants were randomly invited and accepted to participate in a sub-cohort study (MDC cardiovascular cohort). A total of 5533 accepted to participate and attended the second visit at the study screening centre and had overnight fasting venous blood samples drawn for plasma separation. Participants had a number of test measurements done in their blood sample including; fasting blood glucose, plasma insulin, and erythrocyte glycosylated haemoglobin (Hb1Ac) among others.

We excluded 368 of these subjects from the study due to incomplete clinical data, while another 307 subjects did not have plasma available. The remaining 4865 subjects whose plasma samples were sent for analysis, 123 subjects were further excluded because the analysis of their plasma samples did not pass the internal quality control for the biomarker analysis. The remaining 4742 subjects were followed from baseline examination until first event, we excluded 196 with prevalent type 2 diabetes and other 2 more, who had missing MPO results, the remaining 4544 were considered for analysis. Prevalent type 2 diabetes was defined at baseline as; self-reported according to the questionnaire (had a response ‘Yes’ to the question ‘Do you have diabetes’), use of anti-diabetic medication or basing on records from national and regional registers(19).

Consent

All participants gave written consent at enrolment before any study procedures were done and the original study obtained the ethical approval from the Regional Ethical Review Board, Lund, Sweden (LU 51/90).

Identification of incident diabetes

Our outcome measure was incident type 2 diabetes mellitus and was defined as diagnosis made after date of enrolment during follow-up, from baseline until first diagnosis (that is until December 31, 2014) and were identified through national and regional registers. Type 2 diabetes mellitus cases were identified from the Malmo Hb1Ac Register (MHR) at Clinical chemistry, the Swedish National Diabetes Register (NDR), the Swedish Hospital Discharge Register, the Nationwide Drug Prescription register and the Scania region Diabetes 2000 Register(19). For an incident case to be entered in the NDR and Diabetes 2000 register, it required a physician diagnosis after a thorough assessment following the diagnostic criterion of; taking the participants’ fasting plasma glucose concentration with values of equal to 7.0mmolL-¹ or greater which is identical with fasting whole blood glucose of 6.1mmolL-¹ or more(19).

Furthermore, incident type 2 diabetes mellitus cases were identified through reassessment of the cohort. The selection criteria identifying incident cases was either by considering; (i) responses given in the questionnaire, (ii) use of antidiabetic drugs, (iii) fasting plasma glucose and oral glucose tolerance test (OGTT). The OGTT was done by participants having 30g oral load of glucose and after 120 minutes, had whole blood test for glucose concentration, any value equal to 7.0mmolL-¹ or more was considered to be glucose intolerant(20). At least one independent source of registered type 2 diabetes mellitus was a confirmed case.

Biochemical MPO Analysis

Proseek® Multiplex CVD I 96×96 was the reagent kit that was used to measure 92 CVD - related human protein biomarkers, including MPO, simultaneously in plasma frozen at negative 80oC until analysis. A volume of 45mls of venous blood was collected from study participants at baseline and their ethylenediaminetetra acetic acid (EDTA) - plasma concentration of MPO was measured using Proseek Multiplex method, an immunological technique used in biomarker discovery. This technique uses antibodies like in the Enzyme Linked Immunosorbent Assay (ELISA) and the output results are by real time (RT) polymerised chain reaction (PCR). The Proseek reagents are based on a Proximity Extension Assay technology (PEA), where oligonucleotide labelled antibody probe pairs are allowed to bind to their respective target present in the sample in this case MPO biomarker. A PCR reporter sequence was formed by a proximity dependent DNA polymerization event and was subsequently detected and quantified using real-time PCR. The data output was for relative quantification. Therefore, for MPO the data obtained was not of absolute values (pg/ml) but instead described differences in MPO levels between samples and groups.

Measurements and definition of variables

Participants invited at the first visit and consented, were given a questionnaire that was collected at study visit two and was checked for completeness. It assessed for highest education attained categorised as; elementary, primary and secondary, upper secondary, further education without a degree and university degree. The smoking habit was categorised as; never smoked, former smokers, current smoker (regular or occasional smoker). The leisure-time physical activity index was obtained by use of leisure time activity (LTA) questionnaire which included questions on 17 predefined activities and open ended questions(21). The leisure activity comprised of the sum total of all activities adjusted for intensity and was categorised as low for lower quartile, moderate for quartile 2, high for quartile 3 and very high for upper quartile.

Participants’ weight (kg) was taken with balance-beam scale and their standing up height measured using standiometer in centimetres later converted to meters without shoes by two trained nurses (19,22), with participants putting on light clothing and their supine position blood pressure (mmHg) was measured 10 minutes after resting(23). Waist circumference (cm) was taken at the level of the umbilicus with a tape measure. The body mass index (BMI)(kg/m² ) was computed using the standard formula of weight (kg)/square of height(m). Plasma insulin was measured using nonspecific radioimmunoassay(23) set at detection limit of 3mlU/L with intra and inter-assay coefficients of variations 5 and 8% respectively and blood glucose was measured using hexokinase technique(21). Erythrocyte HbA1c was determined by ion exchange chromatography technique, using the Swedish Mono-S standardization with reference values of 3.9 - 5.3 % for non-diabetics(19).

Homeostasis model assessment (HOMA) a measure of insulin resistance was according to the European Group of insulin study of insulin resistance (EGIR) and was determined by the model; [fasting insulin (mlU/L) x fasting blood glucose]/22.5(23).

Statistical Analysis

Statistical analysis was done using IBM - SSPS package, version 22.0 for Windows. Normal distribution of MPO was done using frequency distribution across the sample and reported using a histogram and plasma insulin was widely dispersed and transformed logarithmically for statistical analysis. We used two sided tests with statistical levels of significance assumed when P<0.05. We categorized MPO concentrations into quartiles and calculated descriptive statistics of the sample including mean and standard deviation. The association between MPO quartiles and continuous variables (age, BMI, plasma glucose, plasma insulin, HOMA-IR and HbA1c) were analysed using general linear model. Crude association between MPO quartiles and categorical variables (sex, smoking, education level and LTPA) were analysed using chi square test (table 1).

Cox regression was used to calculate hazard ratios at 95% confidence intervals (CI) to investigate association of MPO quartiles and incident type 2 diabetes mellitus with time of follow-up as the time variable. In the basic model we adjusted for age and sex. In a multivariate model, we adjusted for; age, sex, smoking and education. In another multivariate model, adjusted for; age, sex, smoking, education and BMI. We also performed a model adjusted for age, sex, smoking, education, BMI and HbA1c.

RESULTS

The total sample comprised of 4544 participants and their plasma was analysed for MPO inflammatory biomarker and were composed of 2756(60.7%) females and 1788(39.3%) males. Figure 1 a histogram showing a frequency distribution of MPO across the population. The mean MPO concentration was 3.46(standard deviation of 0.35). Table 1 describes the participant characteristics presented in the MPO four quartiles. It shows that age, BMI, plasma glucose, plasma insulin, HOMA-IR and Hb1Ac all have an increasing trend across the MPO quartiles from Quartile 1 to Quartile 4. MPO concentration decreases with increasing education levels; the majority in the 4th quartile of MPO had elementary education 555(27.0%) as compared with those with university education 76 (22.4%) in the same quartile. Table 2 shows the distribution of incident type 2 diabetes across the four MPO quartiles with the largest number of 228 cases in quartile 4 with the shortest person years of 20083 as compared with the rest.

[...]