Which insulin resistance-based definition of metabolic syndrome has superior diagnostic value in detection of poor health-related quality of life? Cross-sectional findings from Tehran Lipid and Glucose Study

Background The superiority of the diagnostic power of different definitions of metabolic syndrome (MetS) in detecting objective and subjective cardiovascular outcomes is under debate. We sought to compare diagnostic values of different insulin resistance (IR)-based definitions of MetS in detecting poor health-related quality of life (HRQoL) in a large sample of Tehranian adults. Methods This cross-sectional study conducted within the framework of the Tehran Lipid and Glucose Study on a total sample of 742 individuals, aged ≥ 20 years. Metabolic syndrome was defined according to the World Health Organization (WHO), the European Group for the study of Insulin Resistance (EGIR), and the American Association of Clinical Endocrinology (AACE). Health-related quality of life was assessed using the Short Form Health Survey (SF-36). Logistic regression analysis and Receiver Operating Characteristic (ROC) curve were used to investigate the impact of the three IR-based definitions of MetS on HRQoL and compare their discriminative powers in predicting poor HRQoL. Results Compared with other definitions, the WHO definition identified more participants with MetS (41.8 %). Although the AACE definition had higher adjusted odds ratios for reporting poor physical HRQoL (OR: 1.95; CI: 0.84–4.53 and OR: 1.01; CI: 0.55–1.85 in men and women respectively) and mental HRQoL (OR: 0.97; CI: 0.41–2.28 and OR: 1.00; CI: 0.56–1.79 in men and women respectively), none of the three studied definitions were significantly associated with poor physical or mental HRQoL in either gender; nor did ROC curves show any significant difference in the discriminative powers of IR-based definitions in detecting poor HRQoL in either gender. Conclusions None of the three studied IR-based definitions of MetS could significantly detect poor HRQoL in the physical or mental domains, indicating no significant superior diagnostic value for any of these definitions.


Background
Metabolic syndrome (MetS), an escalating health issue worldwide, is a constellation of metabolic abnormalities, which are the major risk factors for developing cardiovascular disease (CVD) and diabetes type 2 [1]. Fast increasing evidence shows the negative effect of MetS on health-related quality of life (HRQoL) as a subjective patient-centered health measurement that concentrates on the individual's own perception of their health status and life satisfaction [2]. Recent data reveals that the prevalence of the syndrome is fast increasing in different populations [3] and is reported to be approximately 34.7 % in Iranian adults [4].
Despite much research have been conducted in recent years, there is uncertainty regarding the concept of MetS and critical investigations have questioned whether the syndrome is a mere aggregation of metabolic abnormalities or a syndrome representing a clinical entity [5]. This doubt has resulted in the introduction of several definitions that include risk factor components which are not entirely similar and more importantly, the cutoffs used for defining them differ. Insulin resistance (IR); the most accepted pathophysiology of MetS, is likely a significant link between the components of this syndrome [1]. On the other hand, obesity, which is included in most definitions of MetS, is identified as the most important correlate of the increasing prevalence of MetS [1]; in addition it has been proposed that IR is significantly related to central obesity [6]. Among the proposed definitions, the World Health Organization (WHO) [7], the European Group for the study of Insulin Resistance (EGIR) [8], and the American Association of Clinical Endocrinology (AACE) [9] emphasize IR as the major component of MetS; however, the National Cholesterol Education Program Adult Treatment Panel III (NCEP-ATP III), the American Heart Association/National Heart, Lung, and Blood Institute (AHA/NHLBI), the International Diabetes Federation (IDF) and the joint interim statement (JIS), emphasize waist circumference (WC) [10].
Considering the above mentioned ambiguity, several efforts have been made to explore the superior diagnostic powers of the WC-and IR-based definitions of MetS in detecting objective and subjective cardiovascular outcomes, i.e. CVD and HRQoL. Regarding objective CVD complications, evidence shows that except for the EGIR definition, the NCEP/AHA, AACE, IDF and modified WHO definitions are all associated with cardiovascular events among an elderly American population [11]. The results of a recent study of the Tehran lipid and glucose study (TLGS) population revealed no differences among diagnostic values of WC-based definitions of MetS in detecting coronary heart disease (CHD) and CVD [12]. Furthermore, in a Dutch population, compared to the WC-based definitions of MetS, IR-based definitions had lower hazards ratios in detecting cardiovascular events [13]. In the field of subjective outcomes, the association between WC-based definitions of MetS and HRQoL, especially in the physical domain and mainly in women has been reported [14]. We recently showed that these definitions failed to show any superiority in the discriminative powers in detecting poor HRQoL in Tehranian adults without diabetes [15]. Although the association between IR and HRQoL has been documented [16] there is no study reporting the association between IR-based definitions of MetS and HRQoL. As one of the first efforts, this study aimed to investigate diagnostic values of three IR-based definitions of MetS including WHO, EGIR and AACE to detect poor HRQoL in a large population of Tehranian adults.

Subjects and design
The current study was conducted within the framework of the TLGS, a large scale ongoing community based prospective study being performed on a representative sample of residents of district-13 of Tehran, the capital of Iran. Details of the rationale and design of the TLGS have been published elsewhere [17]. The TLGS has two major components: 1) a cross-sectional prevalence study of non-communicable diseases and their associated risk factors implemented from 1999 to 2001 and 2) a prospective follow-up study in which non-communicable diseases risk factors are measured approximately every 3 years.
In the current study, a total of 742 individuals, aged ≥ 20 years, with insulin measurements, participating in the TLGS between September 2005 and September 2007 (the second follow-up), was recruited. Information data of these 742 participants was analyzed using the WHO definition. For EGIR and AACE definitions, after excluding 135 persons with diabetes and 17 persons with missing data, the information of 590 persons met the inclusion criteria and was analyzed. The study was approved by the ethics committee of the Research Institute for Endocrine Sciences, Shahid Beheshti University of Medical Sciences and written informed consent was obtained from all participants.

HRQoL measures
To assess HRQoL, we used the Iranian version of Short Form Health survey (SF-36), which has been validated in Iran [18]; this widely used questionnaire contains 36 questions summarized into eight subscales; four physical health related subscales including physical functioning, role limitations due to physical health problems, bodily pain, general health and also four mental health-related subscales including vitality, social functioning, role limitations due to emotional problems, and mental health. The physical subscales are summarized and termed as the physical component summary (PCS) and similarly the four mental subscales are termed as the mental component summary (MCS) [19]. The score attributed to each subscale ranges from 0 to 100 as the worst and the best conditions of health respectively. Calculating of the PCS and the MCS scores was done using the Quality Metric Health Outcomes Scoring Software 2 [20].

Definitions
Metabolic syndrome was defined according to the WHO [7], EGIR [8] and AACE [9] criteria (Table 1). Based on the criteria of the American Diabetes Association (ADA), diabetes was defined as fasting plasma glucose ≥ 7 mmol/ L, 2-h post 75 g glucose load ≥ 11.1 mmol/L or receiving antidiabetic medications [21]. Insulin resistance was defined as the upper quartile of insulin level of baseline population from the second follow-up of TLGS. After excluding patients with diabetes, those who had received educational intervention, and participants of the current study, the cut point for IR was calculated. The median [range] insulin concentration was 7.45 mU/ml [0.2-51.84] with an upper quartile range of ≥10.63. Menopause was defined as the time of cessation of menstrual periods for 12 consecutive months, not due to surgery or any other biological or physiological causes [22]. Impaired glucose regulation was defined according to the criteria of the ADA as fasting blood glucose 5.6 mmol/L to 6.9 mmol/L or 2-h post 75 g glucose load 7.8 mmol/ L to 11.1 mmol/L [21]. Smoking status was considered in two groups: 1) Non-and ex-smokers and 2) Current smokers. Additional information regarding age, physical activity [23] and current use of oral hypoglycemic agents, lipid-lowering and anti-hypertensive medication were obtained using the TLGS data.

Other measures
Waist circumference was measured at the umbilical level, over light clothing, using an unstretched tape meter, without any pressure to body surface and measurements were recorded to the nearest 0.1 cm. Blood pressure was measured twice, after participants were seated for 15 min, using a standard mercury sphygmomanometer; there was at least 30s interval between these two separate measurements and the mean of two measurements was recorded as the participants blood pressure. Twelve-hour fasting blood samples were collected in tubes containing 0.1 % EDTA and were centrifuged at 4°C and 500 × g for 10 min, to separate the plasma. Blood glucose was measured on the day of blood collection by an enzymatic colorimetric method using glucose oxidase. Fasting serum insulin was determined by the electrochemiluminescence immunoassay (ECLIA) method, using Roche Diagnostic kits and the Roche/ HitachiCobas e-411 analyzer (GmbH, Mannheim, Germany). Serum total cholesterol and triglyceride concentrations were measured with commercially available enzymatic reagents (Pars Azmoon, Tehran, Iran) adapted to a selectraautoanalyzer. High density lipoprotein-cholesterol (HDL-C) was measured after precipitation of the apolipoprotein B-containing  [17].

Statistical analysis
By use of graphical methods, continuous variables were checked for normality and are expressed as mean ± SD. Distribution of variables with normal and non -normal distributions between two groups and also categorical variables were compared using sample t-test, Mann-Whitney test and χ2 test respectively; categorical variables are reported as percentages.
Poor HRQoL was defined as the first tertile of PCS or MCS and to compute the odds ratios (ORs), logistic regression analysis was used. Sex specific ORs with 95 % confidence intervals were computed for men and women separately; model 2 was adjusted for age (years) and model 3 was adjusted for age, smoking (only in men-Ref: never smoked or ex-smoker), education (Ref: above high school), menopause (only in women-Ref: productive age) and marital status (Ref: married). In women, smoking was not adjusted in model 3 because

Results
General metabolic and clinical characteristics of study participants are listed in Table 2; compared to women, men had higher mean levels of WC, SBP, DBP (P < 0.001), but lower HDL-C (P < 0.001); compared to women, rates of higher education and married subjects were significantly higher in men (P <0.001). According to different MetS definitions, more participants (41.8 %) met the WHO definition, followed by AACE (30.7 %) and EGIR (25.6 %) criteria. Both in the physical and mental domains, more men had higher scores in all subscales of SF-36 indicating better HRQoL; however the score of role emotional subscale was higher in women by all three definitions. Among the subscales, physical and social functioning got the highest scores in both men and women respectively according to all three definitions (Table 3). Table 4 shows the risk of being in the lowest tertile of PCS and MSC according to MetS status by each three definitions. There was no remarkable difference between the rates of poor HRQoL in any of the definitions or either gender. Unadjusted odds ratios (95 % CI) for poor physical health for the WHO, EGIR and AACE definitions were 1.89 (1.00-3.57), 2.04 (0.83-5.03), 1.75 (0.82-3.71) for men respectively, while, in women, they were 1.51 (0.97-2.37), 1.21 (0.67-2.19), 1.56 (0.92-2.64) respectively, significant only in men based on the WHO definition (Table 4).
Compared with women, after adjustment for confounding variables, in physical HRQoL all definitions showed higher odds ratios in men as follows: WHO (Table 4).
After adjustment for confounding variables, ROC analysis showed no significant superiority in the discriminatory powers of the three different MetS definitions in detecting poor HRQoL in physical or mental domains in either gender. However, especially in physical health, women showed higher AUCs for all definitions, in comparison to men (Table 4).

Discussion
In the current study, after adjusting confounding variables, none of the three IR-based definitions of MetS could significantly detect poor HRQoL in the physical or mental domains. Moreover ROC analysis showed no significant superior diagnostic value for any of the three definitions; however the WHO definition identified more patients with MetS (41.8 %) than the other IR-based  Although previous studies have documented the predictive value of different WC-based definitions of MetS on the objective and subjective outcomes of CVD and HRQoL [13,15], to the best of our knowledge, this is the first report comparing the diagnostic values of the IR-based definitions of this syndrome in detecting poor HRQoL. Findings from studies investigating the impact of MetS on the risk of CVD, as a measurable outcome, show both similarities and differences. Based on previous findings, compared to the IR-based definitions, the WC-based definitions of MetS showed a superior predictive value of non-fatal CVD among an elderly Dutch population [13]; Consistent to this, another study showed that WHO and EGIR definitions were associated with lower risk of all cause CVD than WC-based definitions in men enrolled in a multiethnic study [24]. Moreover, a study conducted in Turkey, reported that  [25].
Although in our previous study, the association between poor physical HRQoL and different WC-based definitions of MetS was significant in women but not in men [14], in the current study it is interesting to note that the odds of having poor physical HRQoL is lower in women than in men, a sex specific difference which may be associated to the use of IR as the main component of MetS definitions used in this study; in this regard, a study by Schlotz et al. assessing the relation between IR and HRQoL, showed a significant association between poor HRQoL and some subscales of PCS in men but not women after controlling for confounding variables [16]. Based on the Masharani et al. findings, this sex specific difference in the association between IR-based definitions of MetS with HRQoL may be due to sex differences in the associations between insulin resistance, regional adipose stores, and lipids values which may result in higher prevalence of obesity and high WC in men in association with IR [27]; this is while previous studies revealed that abdominal obesity is the major component of association between MetS and poor HRQoL in women [28]. On the other hand, in type 2 diabetic patients, which is one of the main manifestations of IR, women compared to men, showed worse quality of life and mental well-being [29].
Based on our findings, the IR-based definitions of MetS were not associated with the mental HRQoL in any of the three definitions, results consistent with those of our previous study in which WC-based definitions of MetS showed no significant association with poor mental HRQoL in adult Tehranians [15]. Moreover previous studies have shown that IR and its related measures are associated with poor HRQoL in the domains of physical health but not in domains of mental health [16]. The association between IR and depression has been shown in some previous studies, including a recent meta-analysis, in which a small, but significant association was observed between IR and depression [30], an observation not found in our study.
Based on our knowledge, this is the first report comparing the diagnostic impact of different IR based definitions of the metabolic syndrome in a large sample of adults in the general population. Our study has some limitations; first, this study was conducted using a cross -sectional design, so we are unable to draw conclusions regarding the causal association between MetS and HRQoL. Second, there may be yet other confounding factors such as depression and economic status that affect HRQoL, factors that we did not adjust for. Third, since the three definitions of MetS tended to be associated with poor physical HRQoL in men, the lack of significant statistical results could be related to low statistical power. Moreover, microalbuminuria, which is included in the WHO definition, was omitted from our study because of lack of data. Finally in the AACE definition, due to inadequate information, we did not include acanthosis nigricans, polycystic ovary syndrome, nonalcoholic fatty liver disease, non-Caucasian ethnicity and sedentary life style in the criteria needed for an individual to be considered at high risk of being insulin resistant.

Conclusions
Although the AACE definition had a higher odds of having poor physical HRQoL compared to the WHO and EGIR definitions of MetS, none the three definitions could detect poor HRQoL after adjusting potential confounders in either physical or mental domains in men and women. Accordingly, ROC analysis failed to show any significant superiority in the discriminatory power of the different definitions over each other.