Epidemiology of metabolic health
Slagter, Sandra Nicole
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2017
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Slagter, S. N. (2017). Epidemiology of metabolic health: Lifestyle determinants and health-related quality of life. Rijksuniversiteit Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
8
Hoofdstukpagina-letteromtrek.indd 15 16/11/16 11:05
8
Summary and general discussion
8
Hoofdstukpagina-letteromtrek.indd 15 16/11/16 11:05
8
8
“May you enjoy the horn of plenty without blowing it.” Bill Copeland
The obesity epidemic in developed countries, while the genetic background has not changed, is due to our obesogenic environment [1]. Everywhere around us there are a lot of responsible factors that provoke unhealthy lifestyle choices, such as physical inactiv-ity and increased consumption of high-calorie foods [2]. Visceral obesinactiv-ity does, however, increase the risk of metabolic complications. Approximately one in four Europeans have the metabolic syndrome (MetS), which places them at higher risk for several chronic diseases, with the most common type 2 diabetes (T2D) and cardiovascular disease (CVD) [3]. MetS even threatens developing countries where the traditional lifestyle is replaced by a more Western-like lifestyle [4]. However, it appears that not all obese individuals develop metabolic complications, despite the fact that they are exposed to the same obesity-promoting environment. This is the so-called metabolically healthy obese [5].
The research described in this thesis aimed to provide an update on the prevalence of MetS and metabolically healthy obesity (MHO), to contribute to a better understand-ing of the associations between lifestyle factors and metabolic health, and in addition, to examine which aspects of health-related quality of life are influenced by obesity and metabolic health complications. To this end, we used data from the large Dutch LifeLines Cohort Study.
variatioN iN thE prEvalENcE of mEtS aNd mho
The prevalence of obesity has continued to rise the last decades in European countries [6]. In our collaborative Healthy Obesity Project (HOP), among ten large European population-based cohort studies (chapter 2), prevalence of obesity varied between 11.6% in Italy to 26.3% in Germany. The earliest data came from 1995 and the most re-cent data from 2012. We also calculated the prevalence of MetS and MHO among obese individuals. Such an extensive comparison has never been done before. We used the well-established clinical risk factors associated with MetS to identify the metabolically healthy obese. MHO was defined as the presence of a BMI ≥30 kg/m2, none of the MetS
components except for waist circumference, and no previous diagnosis of CVD. MetS prevalence increased with ageing, while the MHO prevalence decreased with ageing. When corrected for age, MetS was more common among men (43-78%, 56% in Life-Lines) than among women (24-65%, 37% in LifeLife-Lines), while MHO was more frequent in women (7-28%, 23% in LifeLines) compared to men (2-19%, 10% in LifeLines). Our re-sults demonstrated a significant diversity in the prevalence of MetS and MHO across the European cohort studies, despite the use of a uniform definition and age-standardized prevalence estimates. Although, we obtained our data from large population-based
cohort studies, the results cannot always be generalized to the overall prevalence in the specific countries, as some cohorts have only collected data from a specific region (Italy, Norway and the Dutch PREVEND study), or from a specific age group (United Kingdom).
Another finding of the study was that elevated blood pressure was the most fre-quently occurring factor contributing to the prevalence of MetS. This has also been confirmed in the Dutch normal weight and overweight population, where high blood pressure and increased waist circumference were most frequently present components of MetS (chapter 7). MetS was not only seen among obese (51%) and overweight individuals (19%), but also in a small subset of the normal weight population (4%) (Figure 1A). Within all these BMI groups, MetS was more common among men. There are several explanations for this observation. First, men have more elevated visceral and hepatic fat, making them more susceptible to insulin resistance and the development of other MetS features, while women do have more total body fat [7]. Second, elevated blood pressure is already often present among young men. They will, therefore, meet the criteria for MetS sooner than women. Last, there are studies suggesting that the female hormone estrogen may protect against the development of cardiometabolic risk factors [8-10]. This may explain why after the age of 50 years, when in general the menopause starts, the difference in MetS prevalence between men and women becomes smaller (Figure 1B). We found that in the entire population prevalence of MetS increased with age (Figure 1B), which can be explained by the age-related increase in blood pressure, waist circumference and fasting glucose (chapter 7).
The detected variation in metabolic health prevalence between countries, suggest that on top of age, sex and genes our environment is accountable [11, 12]. Indeed, a high
0 10 20 30 40 50 60 70 Normal
weight Overweight Obese
Pr ev ale nce , % BMI class Men Women A 0 10 20 30 40 50 60 70 18-29 30-39 40-49 50-59 60-69 70-79 Pr ev ale nce , % Age group Men Women B
8
blood pressure (to some extent, depending on age) and waist circumference can bemodified by adapting a more healthy lifestyle. Therefore, targeting lifestyle may play an important role in reducing the development of MetS as well as becoming or remaining metabolically healthy. However, first we need to understand how lifestyle is associated with MetS and its individual components.
thE iNfluENcE of lifEStylE oN mEtS
There is extensive evidence that many health benefits can be gained through favour-able lifestyle modifications. Smoking, too much alcohol, poor diet and lack of physical activity are common risk factors. It has been estimated that approximately 50% of cases of coronary heart disease, stroke and diabetes are attributable to these lifestyle factors [13, 14]. In the following section the results of the studies in this thesis regarding lifestyle factors will be summarized and discussed.
Smoking is bad for lipids
In the Netherlands, an important risk factor for health problems is still smoking, despite the fact that the prevalence of smoking decreased from 60.0% in 1958 to 18.4% in 2012 [15]. It has been suggested by several small-scale studies [16-19] and one meta-analysis [20] that smoking contributes to MetS. In this meta-analysis data from 13 prospective studies, involving 56,000 individuals, a positive association was found between smoking and risk of MetS, although the increased risk was only significant for heavy smokers and not among light and former smokers [20]. Our study in chapter 3 is the first very large study, including almost 60,000 individuals, supporting the association between smok-ing and MetS in both men and women, in all BMI classes, and in a dose-related manner. In the LifeLines study population, 21.3% of participants were current smokers. Light-, moderate- and heavy smokers had lower levels of HDL cholesterol (HDL-C), higher levels of triglycerides, and a larger waist circumference compared to non-smokers. While this relationship was found for both men and women, irrespective of their BMI, there was no clear association between smoking and either blood pressure or fasting blood glucose. Furthermore, the amount of smoking was also associated with unfavourable changes in the levels of apoA1 and apoB and the HDL and LDL particle size. The latter finding may provide a new pathophysiological mechanism that links smoking to increased risk of CVD.
There are some biologically plausible explanations linking smoking to increased risk of MetS. Smoking leads to acute and chronic changes in the balance of the autonomic nervous system, resulting in sympathetic predominance, which increase the risk of cardiovascular events [21, 22]. Also, smoking increase the levels of circulating
insulin-antagonistic hormones levels, such as cortisol and growth hormone [22]. As a result the glucose and lipid metabolism is affected. Although in our study fasting blood glucose levels were only marginally increased in smokers, there are some studies showing that compared to non-smokers, active smokers have more serious insulin resistance and hyperinsulinaemia, which increase their risk for T2D [23]. Furthermore, data exist that smokers (especially heavy smokers) have a higher BMI than non-smokers [24] and great-er risk of abdominal fat accumulation [25]. The increased plasma cortisol concentrations seen in smokers are in part responsible for the accumulation of visceral fat, which, in turn, increases waist circumference [26].
While our study suggests that smoking may unfavourably change certain MetS com-ponents, it can be debated whether smoking by itself may increase the prevalence of MetS, or whether it is related to other unhealthy lifestyle factors. For instance, smoking is associated with higher alcohol consumption, lower consumption of fruit and vegetables, and less leisure time physical activity [27]. Therefore, the next step was to evaluate the effect of the combined use of tobacco and alcohol on MetS.
light alcohol consumption may partly counteract effects of smoking
Compared to other European countries, the consumption of alcohol is rather moderate in The Netherlands. Dutch people consumed an average of 9.9 liters of pure alcohol per capita (15+ years of age) per year in the period 2008-2010, while the European average is 10.9 liters a year1. However, compared to 50 years ago alcohol consumption
has increased in the Netherlands, and since 1990 the level of alcohol consumption has remained stable1. Moderate alcohol consumption has consistently been associated with
a decreased risk of T2D and CVD in prospective cohort studies compared with absten-tion or excessive consumpabsten-tion [28, 29].
Alcohol consumption often clusters together with smoking [30]. In chapter 4 we carefully assessed the combined effects of smoking and alcohol consumption on MetS and its individual components. MetS was least prevalent among subjects with light to moderate alcohol consumption (≤1-2 drinks/day), in normal-weight and overweight subjects, irrespective of their smoking status. Among obese former and current smok-ers, non-drinkers or those with light alcohol consumption (≤1 drink/day) showed the lowest MetS prevalence. In figure 2 the observed associations of smoking and light alcohol consumption (relative to not drinking) with the individual MetS components are summarized.
We found that especially light alcohol consumption was associated with a favour-able effect or no effect on the individual MetS components. Light alcohol consumption might therefore partly compensate for the unfavourable associations of smoking with
8
MetS. An interesting finding of chapter 3 is that in people with normal weight, heavyalcohol consumption showed a trend towards a larger waist circumference in non-smok-ers and light-moderate smoknon-smok-ers. In contrast, light-moderate alcohol consumption was associated with a smaller waist circumference in overweight non-smokers and former smokers, and obese non-, former and heavy smokers. In chapter 2 we did, however, report that smokers had a larger waist circumference, especially heavy smoking obese women. A possible explanation for the results described in chapter 3, is that in general light-moderate drinking, especially wine consumption, is associated with healthier over-all dietary and lifestyle choices, which may have led to less abdominal obesity [31, 32].
The most popular alcoholic drink in the Netherlands is beer, which accounts for more than half of all alcohol consumed. Next is wine with 36% and spirits with 17%1. We found
that wine drinking was characterized by an overall better metabolic profile and showed a protective association for MetS, compared to that of non-drinkers and drinkers of mainly beer or spirits (chapter 3). Both alcohol and non-alcohol components of wine, such as polyphenols, may be responsible for the lower prevalence of MetS and some of the MetS components. In general, alcohol consumption may increase HDL-cholesterol and reduce insulin resistance [33, 34]. However, non-alcohol components of wine have also been shown to increase HDL-cholesterol and lower triglycerides [35, 36]. The healthier lifestyle of wine consumers is an important factor as well [31, 32]. Sluik et al. suggested that alcoholic beverage preference is merely a proxy for socio-demographic and other lifestyle factors, rather than independently related to health status [32].
Occasional alcohol drinking is well accepted as a favourable lifestyle factor for car-diovascular health. Although we would not encourage alcohol consumption to
abstain-1 http://www.who.int/substance_abuse/publications/global_alcohol_report/profiles/nld.pdf + = a favourable association, - = an unfavourable association, N = neutral.
+/-+ ++ + N N N + -- HDL-cholesterol Triglycerides Waist circumference Blood pressure Blood glucose
-figure 2. Overview of the associations of smoking and light alcohol consumption (relative to no alcohol
ers, our results suggest that light alcohol consumption will not negatively influence the development of MetS. Smoking cessation and reduction of excessive alcohol consump-tion seem promising steps in the management of MetS, and prevenconsump-tion of its sequelae. food choices and physical activity may improve metabolic health
It is well known, that the first step in improving metabolic health is weight reduction. The best way to lose the extra pounds is by restricting energy intake and increasing physical activity. In practice, weight loss by energy-restricted diets are only successful for a few months [37]. Only one in six overweight and obese adults reported to ever have maintained weight loss of at least 10% for one year in the National Health and Nutrition Examination Survey (1999-2006) [38]. One of the reasons for the poor long-term out-come of weight-loss diets is that the dietary advises are too different from the existing patterns of food consumption in a (sub-)population. Adherence to the strict regimens of the diet is therefore difficult and the motivation typically diminishes with time.
As described in chapter 2, there is a subgroup of obese individuals that, despite their excessive weight, has not (yet) developed any metabolic abnormalities except abdominal obesity. Between countries there is still a diversity in the reported prevalence of obese subjects with MetS and MHO, when taking age and sex into account (chapter
2). This have led to the hypothesis that independently of obesity, modifiable factors,
such as diet and physical activity, still contribute to the differences in metabolic health. In chapter 5, we assessed obesity-specific dietary patterns using principal com-ponent analysis (e.g. a form of factor analysis), and compared these dietary patterns and physical activity between MHO and metabolically unhealthy obesity (MUO). While moderate alcohol consumption and not smoking were positively associated with MHO, in both men and women, we observed sex-specific differences in dietary patterns and physical activity associated with MHO.
Our data showed that only two of the four obesity-specific dietary patterns were associated with metabolic health in women only. A higher score on the ‘fruit, vegetables
and fish’ pattern, which reflected high consumption of mainly fruit, vegetables, fish and
fermented milk products, was positively associated with MHO. The ‘bread, potatoes and
sweet snacks’ pattern reflected a diet of mainly foods with a high glycaemic index (GI),
e.g. bread, potatoes, sweet sandwich toppings, pastries, biscuits and desserts. A higher score of this pattern was inversely associated with MHO. These associations were not found for men. One possible explanation for this may be that women reported higher consumption of the specific types of foods that contributed high to the ‘fruit, vegetables
and fish’ pattern compared to men. Also pastries, biscuits and desserts were foods more
consumed by women. We also suggested that men and women may have a different view on improving health status. Women might prefer dietary changes to lose weight, while men might prefer changes in physical activity to improve their fitness.
8
Only one recent study from Australia used the factorial method, i.e. principalcom-ponent analyses, to link dietary patterns to MHO [39]. The authors found that a higher score on the ‘healthy’ dietary pattern (with high intakes of whole grains, fresh fruit, dried fruit, legumes and low fat dairy) was associated with a more healthy metabolic profile. In contrast to our study, the authors derived dietary patterns from a population includ-ing obese and non-obese people instead of obesity-specific patterns. Furthermore, the authors used different criteria to define MHO than used in our study [39]. Still, both the results of the Australian and our study increase the confidence that dietary pattern analysis can be meaningful in understanding the metabolic health.
Metabolically healthy obese men were characterized by higher engagement in intensive vigorous physical activity, while among women there was no association of physical activity with MHO. The reason for this finding needs to be explored further. We did, however, hypothesize that the higher level of physical activity in men compared to women, may have strengthened the association between vigorous physical activity and MHO. Furthermore, it is possible that women experience and, therefore, report certain activities as more intensive than men, but actually a lower cardiorespiratory response is obtained. Hence, as a result men may develop a higher cardiorespiratory fitness level than women, which has been linked to MHO [40].
The results described in chapter 5 generates the hypothesis that the preferred MHO phenotype may be maintained or even transition from unhealthy to healthy obesity can be achieved through diet and physical activity, without aiming for weight loss per se. However, we should keep in mind that lifestyle factors are interrelated and no exclusive lifestyle factor could independently affect health status.
health-related quality of life compromised in obesity
Research often focuses on the long-term effects of obesity on health and mortality. While people overestimate health risks which they cannot (or hardly) influence, people have in general an optimistic view about lifestyle-related health risks [41, 42]. Adaption of life-style changes are more successful if there is a reward within a short period of time [43]. To motivate obese individuals to change their lifestyle, therapeutic strategies should, therefore, also consider (existing) decrements in the physical and mental health of the individual. In chapter 6 we assessed with the RAND 36-Item Health Survey (RAND-36) [44], which domains of health-related quality of life (HR-QoL) were affected by obesity together with obesity-related conditions, e.g. T2D, MetS and inflammation.
Previously, the association between obesity and HR-QoL has been investigated in a variety of settings, including centers for weight loss, general medical practices, and the general population [45-50]. Physical, and to a lesser extent, psychosocial impairments were found to be more severe in obese women than in men [47, 49-51]. In chapter 6, we showed that low scores on several domains of HR-QoL were enhanced by the grade
of obesity, T2D, MetS, and level of inflammation. Significant associations were found between these conditions and lower HR-QoL scores related to general health and physi-cal functioning in both obese men and women. However, in men, obesity grade and T2D were associated with more frequent impairment of physical functioning than in women. In contrast, inflammation level assessed by high sensitive C-reactive protein measurements was associated with more frequent impairment of physical functioning in women. These results suggest that in men and women, the presence of morbidity may operate differently in the pathway linking obesity to reduced HR-QoL.
In both men and women, mentally-oriented domains (especially vitality and social functioning) were marginally more often impaired among subjects with a higher obesity grade, T2D, MetS or with higher levels of inflammation. Although the RAND-36 does in-clude questions regarding pleasant (happy, lot of energy, full of pep, peaceful and calm) and unpleasant feelings (worn out, nervous, tired, downhearted and blue, felt so down that nothing could cheer you up), the questions are not necessarily aimed at psycho-logical problems often present in obese individuals. Such psychopsycho-logical problems might include depression, eating disorders, distorted body image, and a low self-esteem [52].
The negative associations found between obesity and obesity-related conditions and HR-QoL suggest that obese people already experience adverse effects of obesity before they develop more severe conditions such as CVD. Assessment of HR-QoL in obese individuals can help caregivers to offer a more personalized approach in obesity treatment. Placing specific attention to the affected domains of HR-QoL may help in setting personal goals.
thE uSE of mEtS aNd mho iN EpidEmiology
The prevalence of MetS and MHO, as well as their individual components, are entirely dependent on the choice of cut-off values used to define an impaired metabolic profile. In the existing definition of MetS and MHO a dichotomization has been made at two levels: 1) first, every single risk factor is dichotomized using a threshold; and 2) then, the diagnosis of MetS or MHO is made if a certain number of risk factors is present or absent. used definition for metS and mho
Throughout this thesis, MetS was defined according to the revised version of the National Cholesterol Education Program’s Adult Treatment Panel III (NCEP ATPIII) [53]. Compared to the original NCEP ATPIII, not only treatment for dyslipidaemia, hyperglycaemia or hy-pertension has been included in the revised version, but also the threshold for impaired fasting glucose has been lowered from ≥6.1 mmol/L to ≥5.6 mmol/L [54]. Although this new threshold for fasting glucose is generally accepted as part of the MetS definition,
8
we chose to use the old threshold of ≥6.1 mmol/L in the definition for MHO. The lenientthreshold for impaired fasting glucose has the potential to label one third of the general population as ‘high risk’, while the majority of them will never develop diabetes [55]. In addition, subjects with fasting glucose 5.6–6.0 mmol/L do not have an increased risk of CVD in contrast to subjects with a fasting glucose 6.1–6.9 mmol/L [56, 57]. So instead of using the strict cut-off for glucose, we were more strict on the number of MetS components needed to be absent to define MHO. Our definition of MHO was considered as the absence of all MetS components, as defined by the original NCEP ATPIII (with ad-ditionally including treatment for dyslipidaemia, hyperglycaemia or hypertension) [53], a BMI ≥30 kg/m2, and no history of CVD. Waist circumference was excluded from these
criteria because it highly correlates with BMI ≥30 kg/m2 [58]. Defining MHO as having
none of the components may have greater utility as well, to show a greater difference in risk for CVD and T2D.
cut-off points for identification of abdominal obesity
Abdominal obesity is an important feature of insulin resistance and MetS [53]. Waist circumference, as a measure for abdominal obesity, is therefore one variable of MetS. As reported in our study (chapter 7) and previous ones, abdominal obesity is more com-mon acom-mong women than acom-mong men [59-61]. This appears to be in contrast with our knowledge that women have more total body fat and men have more visceral and he-patic fat [7]. The purpose of defining abdominal obesity is to identify those individuals at increased risk for obesity-related cardiomatabolic disease. Because this risk may depend on ethnicity, it was suggested that ethnicity specific cut-off values should be used [62]. There are two cut-offs to define abdominal obesity in Europeans: i.e. the high cut-offs in the widely used NCEP ATPIII definition (≥102 cm in men and ≥88 cm in women), and the lower cut-offs (≥94 cm and ≥80 cm in respectively, men and women).
Since MetS can even develop in non-obese individuals, it is important that the cut-offs for abdominal obesity are set at values that are highly predictive for CVD. However, it is interesting to know that the currently used low- and high levels of abdominal obesity are based on correlations with BMI, respectively ≥25 kg/m2 (overweight) and ≥30 kg/m2
(obesity), instead of resembling an ‘optimal’ point at which CVD risk increases [63, 64]. Moreover, in 2014, a cross-sectional study among German Caucasians showed that even the ≥88 cm (high) cut-off was too low for capturing CVD risk in women, while the ≥94 cm (low) cut-off seemed to be appropriate in men [65].
To our knowledge, no longitudinal studies have assessed the performance of the low- and high cut-offs of abdominal obesity for cardiovascular risk prediction. We be-lieve that re-evaluation of the waist circumference thresholds for Europeans, at which decisions are made relating to health care provision and screening, is still required. Large cohorts with follow-up data such as PREVEND, HUNT and FINRISK are suitable studies
to select the most appropriate waist circumference values for CVD risk prediction. The anticipated long-term follow-up of this LifeLines Cohort Study can undoubtedly add to these prospective assessments.
cut-off points for identification of ‘elevated’ blood pressure
In chapter 2 and 3, we already observed that the prevalence of an elevated blood pres-sure was remarkably high in the population compared to the other MetS components. This is the result of the strict threshold of ≥130/85 mmHg. For individuals with blood pressure levels ranging from 120–139 mmHg systolic and/or 80–89 mmHg diastolic, lifestyle modifications, such as salt restriction, moderation of alcohol consumption, weight reduction and more physical activity, could reduce blood pressure, decrease the progression of blood pressure to hypertensive levels with age, or prevent hypertension entirely [66]. Lifestyle changes often can only control mild elevations of blood pressure and therefore drug therapy might still be necessary. The natural course of increasing blood pressure with ageing has, however, not been taken into account in the definition of MetS. Therefore, more than half of the elderly people (≥60 yrs) will already be as-signed one risk factor of MetS (chapter 7).
The threshold for elevated blood pressure in MetS is different from the blood pressure levels which warrant (medical) treatment according to current international guidelines [67]. In the eight report of the Joint National Committee (JNC 8) age-specific thresholds are advised, endorsing a blood pressure goal of <150/90 mmHg starting at age 60 years and a blood pressure goal of <140/90 mmHg for those below the age of 60 years [68]. In chapter 7, we applied these age-adjusted blood pressure thresholds to emphasize the illogical choice of not taking age into account for the blood pressure component. Accordingly, the ‘elevated’ blood pressure component was 6.0-36.3% less frequent pres-ent in differpres-ent groups of the population, compared to when the strict threshold was used. Especially among men aged <60 years the prevalence of elevated blood pressure decreased when the age-adjusted thresholds were used. Another interesting finding was that, although, the prevalence of elevated blood pressure (with the age-adjusted thresholds) matched better with the prevalence of those treated for hypertension, under-treatment of ‘elevated’ blood pressure was still high among young men.
Compared to the other MetS components, the particular high prevalence of elevated blood pressure in our study population suggests that the threshold of ≥130/85 mmHg is too strict. Even a systolic blood pressure target of <140 mmHg, is too optimistic in the elderly as has been shown by placebo-controlled trials [67]. Although LifeLines is a well-designed cohort study, which collects longitudinal data, at this moment only cross-sectional data are available. We urge other large cohorts with follow-up data to establish better evidence on age-related blood pressure thresholds associated with increased risk for T2D and CVD.
8
the predictive ability of metS and mhoPrevalence statistics of MetS are useful in providing an estimate of the current risk factor burden and the likely burden of CVD and T2D that will result. Yet, it is suggested that the value of MetS beyond that of its individual components or traditional risk factors as a predictor of both all-cause mortality (relative risk ~1.5) and CVD (relative risk ~2.0) is modest at best [69]. However, all available definitions of MetS showed to be a stronger predictor of T2D (estimated relative risk 3.5 to 5.1) [70]. This is likely due to inclusion of components such as fasting glucose and abdominal obesity, which are more strongly as-sociated with diabetes. Unfortunately, MetS does not comprise the full range of clinically valid CVD risk factors. Even the godfather of MetS, Gerald M. Reaven, is concerned about its clinical value, and he bids farewell to this syndrome with the words requiescat in pace (rest in peace) [71]. To improve the predictive ability of MetS for T2D and CVD, other features could be added to the MetS definition, such as C-reactive protein, non-HDL cholesterol or apoB [72, 73], or we could use a more gradual or continuous approach of MetS [74]. This is because dichotomizing continuous variables results in a loss of predic-tive power. Also interventions are seldom started at the levels proposed by the revised NCEP ATPIII. So, individuals will soon be labelled as ‘unhealthy’, but will not receive an intervention either medical or aimed at improving lifestyle. Nowadays in Europe, treat-ment decisions are based on the validated tool the Systematic COronary Risk Evaluation (SCORE) risk chart, estimating the 10-year absolute risk for fatal CVD in patients [75]. For the Netherlands the SCORE risk chart has been calibrated with Dutch CVD mortality data. The scoring system is gender-specific and includes age, smoking status, systolic blood pressure and the ratio total cholesterol/HDL cholesterol. An important target for risk reduction of CVD is blood pressure and low-density lipoprotein (LDL) cholesterol. When a patient has a medium- or high 10-year risk for CVD, LDL-lowering therapy is initiated to reduce LDL cholesterol below 2.5 mmol/L [75]. However, despite achieving the recommended level for LDL cholesterol, many patients retain a high CVD risk. This ‘residual risk’ is mainly due to elevated triglyceride and low HDL-cholesterol levels [76]. It is well known that more risk factors and higher levels of these risk factors increase the absolute CVD risk in a continuous manner. It would therefore be worthwhile to consider an aggregation of the MetS components (i.e. triglycerides, HDL cholesterol, waist cir-cumference and fasting glucose) with SCORE to improve the primary prevention of CVD . As obesity will remain a public health concern for the coming decades, the MHO phenotype will become increasingly important. Meta-analyses have indicated that adults who are metabolically healthy obese have a risk of T2D [77], CVD and mortality [78] that is intermediate between that of healthy normal weight and unhealthy obese adults. However, MHO is a temporary state for a sizeable proportion of obese adult population, and many of these individuals will shift to become metabolically unhealthy obese [79-81]. While ageing has a large responsibility in this, poor health behaviours
(chapter 6) and changes in fat distribution (accumulation of visceral fat) may also play an important role in the progression towards less healthy phenotypes and disease [82]. We must, therefore, be cautious in using the term ‘healthy’ obesity. Examination of the MHO phenotype has gained interest the last years, but many questions remain regard-ing the MHO definition and its determinants, the stability of the condition over time, and the long-term health outcomes (not only CVD and T2D) [83].
mEthodological coNSidEratioNS
In this section methodological issues will be discussed, as they are important for a cor-rect interpretation of our results. The studies described in this thesis are largely based on data from the LifeLines cohort study. Although follow-up data will be collected for the next 30 years, at the time of conducting our studies, only cross-sectional data were avail-able from participants enrolled in the study between 2006-2013. Since cross-sectional data are subject to certain limitations, several general, but iterative aspects will be briefly commented on to put the observed findings into perspective.
cross-sectional design
Cross-sectional means that the derived data comes from a single time point, and therefore one cannot be sure that the development of the outcome of interest (MetS and its individuals components), was preceded by the presence of putative risk factors (lifestyle). This being so, it is not possible to infer causality. Nevertheless, the observed associations in our studies are particularly useful for generating hypotheses and provid-ing directions for future studies [84].
Sample selection
The criteria used to recruit participants and the response rate, determine how well re-sults of the conducted studies can be generalized to the population as a whole. In total, 49% of the included participants in the LifeLines cohort study, i.e. those in the age 25-50 years, were originally invited through their general practitioner, while 38% volunteered to participate via participating family members and 13% self-registered via the LifeLines website. Almost 25% of the invited persons agreed to participate, which was compa-rable to the response rate in several other large-scale population studies [85]. A major concern in population-based cohort studies is selection bias, a systematic error induced if the association between exposure and outcome differs for those participating and those who do not participate [86]. The presence of selection bias in the LifeLines cohort study could, therefore, lead to results only representative for the cohort sampled, rather than for the entire adult Dutch population from the northern part of the Netherlands.
8
Compared with the population of the north of the Netherlands, LifeLines participantswere more often female, middle aged, married, living in a semi-urban place and Dutch native [87]. However, adjusted for differences in demographic composition, LifeLines was found to be broadly representative on socioeconomic characteristics, weight status, smoking, the prevalence of major chronic diseases and general health. Although it is no strict guarantee against selection on other variables, the recruitment strategy had no substantial effect on the representativeness of the LifeLines population [87]. This suggests that risk estimates are likely to represent real associations in the general adult population from the northern part of the Netherlands.
However, we still have to deal with the fact that MetS is a very heterogeneous phe-notype. This makes it more difficult to find associations between the exposure of interest (environmental factors) and MetS if any exist. Subtypes of MetS, i.e. the pre-dominant clustering of separate MetS components, may depend on sex, age and BMI (chapter 7), but may also differ between populations [61]. Therefore, if an association between a feature and MetS has been found in LifeLines, it is possible that it may not be present in a population with a different sex, age and BMI distribution or in a non-Dutch population. Genetic and ethnic background may partially account for the differences seen between countries as well [88, 89]. A way to get around the heterogeneity problem of MetS is by providing results by more detailed subgroups. In our studies we included only subjects with a Western-European descent, and stratified the population sample as much as pos-sible by sex and BMI, adjusting for age. In that way the findings are more group-specific and does not depend on the distribution of sex and BMI in the study population. In addi-tion, there is a better chance of finding potential differences in the association between the exposure and outcome measure.
Data availability
A particular strength of the LifeLines cohort study is the large sample size and bioma-terial collection, the extensive physical examination, and the range of topics covered in the baseline questionnaire (demographics, health status, lifestyle and psychosocial aspects). Despite this, the researcher is strongly dependent on the data that is available at the stage of conducting the study. An limitation of our studies is that the presented results could be in part influenced by other important (lifestyle) factors, that are both associated with the exposure and the outcome but does not lie in the causal pathway, i.e. confounding [86]. Gradually we were able to account for multiple important lifestyle factors: chapter 3 – smoking, chapter 4 – smoking and alcohol, and chapter 5 - nutrition, physical activity, smoking and alcohol. Nevertheless, the findings in this thesis should be interpreted with caution. The overall lifestyle of smokers is generally considered differ-ent from the lifestyle of non-smokers. Smoking is associated with lower levels of physical activity, higher alcohol consumption, and specific eating habits [27]. But then, light or
moderate drinking –especially of wine- is usually associated with a healthier lifestyle [31, 32]. In future studies multiple lifestyle factors should be analysed, including the possibility of multiple interactions between lifestyle factors.
Questionnaires
Data on demographics, health status, lifestyle and psychosocial aspects were assessed by a self-reported questionnaire. The quality of the data therefore, is determined to a large extent on the patient’s ability to accurately recall past exposures. Recall bias may result in either an underestimate or overestimate of the association between exposure and outcome [86]. There are, however, two types of recall bias: differential and non-differential. Differential recall bias means that cases have a different recall than controls, while with non-differential recall bias both cases and controls have the same recall. In prospective cohort studies, non-differential recall bias is more likely to occur, since there are no prior hypotheses regarding the exposure-outcome associations [86, 90]. Hence, individuals do not know if they are a ‘case’. In our studies, most of the participants are not aware that they have a metabolic risk factor or even MetS. Therefore, the risk of a selective memory is reduced. However, behaviours or habits which are socially less desirable such as smoking, alcohol abuse, or overconsumption of foods are more often prone to underestimation, while physical activity is often overestimated [90]. Therefore, we ranked individuals into categories, rather than using the estimated absolute quanti-ties, which reduces the effect of over- or underreporting. Thus, recall bias is expected to have only a minor effect on the presented results in this thesis.
8
coNcluSioN aNd futurE pErSpEctivESIn this thesis we had a closer look at the epidemiology of metabolic health, its associ-ated lifestyle risk factors and the health-relassoci-ated quality of life in obese individuals. We confirmed that, together with (overweight and) obesity, MetS is very common in different European countries and the Netherlands (the Dutch population living in the Northern part). Although not so prevalent, in the Dutch LifeLines cohort still almost 1 out of 4 obese women and 1 out of 10 obese men are metabolically healthy obese, de-pending on their age. Lifestyle factors, such as smoking, alcohol consumption, diet and physical activity, may play an important role in understanding why some people do have metabolic abnormalities, while others remain healthy or develop MetS at a slower pace. However, the obese without (multiple) metabolic dysregulations may still experience a lower health-related quality of life. Therefore, it is recommended to implement assess-ment of health-related quality of life in the treatassess-ment of obesity. Follow-up data and more in-depth measurements of lifestyle factors can optimize the prediction of people who will progress from metabolically healthy to metabolically unhealthy, and whether reversal of the process is possible. This will offer a promising window of opportunity for the prevention of MetS and hence, T2D and CVD.
MetS is still a concept which is especially used in research settings rather than in clinical practice. A famous quote of Albert Einstein is: “Everything should be made as
simple as possible, but not simpler”. Perhaps the current definition of MetS is made too
simple. Prevention can only be effective if we identify persons at real risk for developing MetS and understand the interactive background of metabolic health. To do so, future research may include:
– Improvement of the definition of MetS and MHO, by adding biochemical measure-ments more predictive for CVD.
– Investigating the use of a continuous metabolic risk score, which might be an inte-gration of the MetS components with the SCORE risk chart.
– Identification of metabolic health across racial/ethnic and socioeconomic lines, which may help to set priorities for interventions.
– Examination of lifestyle interactions with metabolic health.
– Gain insight in the role of genes to better understand why some (obese) individuals do not develop or experience delayed development of metabolic abnormalities. – Evaluation of possible gene-environment interactions in metabolic health.
rEfErENcES
1. Bouchard C: The biological predisposition to obesity: beyond the thrifty genotype scenario. International journal of obesity (2005) 2007, 31(9):1337-1339. 2. Wells JC: The evolution of human adiposity and
obesity: where did it all go wrong? Disease models & mechanisms 2012, 5(5):595-607.
3. Grundy SM: Metabolic syndrome pandemic. Arterioscle-rosis, thrombosis, and vascular biology 2008, 28(4):629-636.
4. Kassi E, Pervanidou P, Kaltsas G, Chrousos G: Metabolic syndrome: definitions and controversies. BMC medicine 2011, 9:48.
5. Denis GV, Obin MS: ‘Metabolically healthy obesity’: origins and implications. Molecular aspects of medicine 2013, 34(1):59-70.
6. Finucane MM, Stevens GA, Cowan MJ, Danaei G, Lin JK, Paciorek CJ, Singh GM, Gutierrez HR, Lu Y, Bahalim AN et al: National, regional, and global trends in body-mass index since 1980: systematic analysis of health examination surveys and epidemiological studies with 960 country-years and 9.1 million participants. Lancet (London, England) 2011, 377(9765):557-567. 7. Geer EB, Shen W: Gender differences in insulin
resis-tance, body composition, and energy balance. Gender medicine 2009, 6 Suppl 1:60-75.
8. Cignarella A, Kratz M, Bolego C: Emerging role of estro-gen in the control of cardiometabolic disease. Trends in pharmacological sciences 2010, 31(4):183-189. 9. Munoz J, Derstine A, Gower BA: Fat distribution and
insulin sensitivity in postmenopausal women: influ-ence of hormone replacement. Obesity research 2002, 10(6):424-431.
10. Rosano GM, Vitale C, Mercuro G: The metabolic syn-drome in women. Women’s health (London, England) 2006, 2(6):889-898.
11. Andreassi MG: Metabolic syndrome, diabetes and atherosclerosis: influence of gene-environment inter-action. Mutation research 2009, 667(1-2):35-43. 12. Pataky Z, Bobbioni-Harsch E, Golay A: Open questions
about metabolically normal obesity. International journal of obesity (2005) 2010, 34 Suppl 2:S18-23. 13. Raad voor de Volksgezondheid en Zorg: Preventie van
welvaartsziekten: effectief en efficiënt georganiseerd. In: Den Haag, december. 2011.
14. The four domains of chronic disease prevention 15. OECD: Health at a Glance: Europe 2014. In: OECD
Publishing. 2014.
16. Geslain-Biquez C, Vol S, Tichet J, Caradec A, D’Hour A, Balkau B: The metabolic syndrome in smokers. The D.E.S.I.R. study. Diabetes & metabolism 2003, 29(3):226-234.
17. Weitzman M, Cook S, Auinger P, Florin TA, Daniels S, Nguyen M, Winickoff JP: Tobacco smoke exposure is as-sociated with the metabolic syndrome in adolescents. Circulation 2005, 112(6):862-869.
18. Wilsgaard T, Jacobsen BK: Lifestyle factors and incident metabolic syndrome. The Tromso Study 1979-2001. Diabetes research and clinical practice 2007, 78(2):217-224.
19. Zhu S, St-Onge MP, Heshka S, Heymsfield SB: Lifestyle behaviors associated with lower risk of having the metabolic syndrome. Metabolism: clinical and experi-mental 2004, 53(11):1503-1511.
20. Sun K, Liu J, Ning G: Active smoking and risk of meta-bolic syndrome: a meta-analysis of prospective studies. PloS one 2012, 7(10):e47791.
21. Curtis BM, O’Keefe JH, Jr.: Autonomic tone as a cardiovascular risk factor: the dangers of chronic fight or flight. Mayo Clinic proceedings 2002, 77(1):45-54. 22. Middlekauff HR, Park J, Moheimani RS: Adverse effects
of cigarette and noncigarette smoke exposure on the autonomic nervous system: mechanisms and implica-tions for cardiovascular risk. Journal of the American College of Cardiology 2014, 64(16):1740-1750. 23. Willi C, Bodenmann P, Ghali WA, Faris PD, Cornuz J:
Active smoking and the risk of type 2 diabetes: a systematic review and meta-analysis. JAMA 2007, 298(22):2654-2664.
24. Chiolero A, Jacot-Sadowski I, Faeh D, Paccaud F, Cornuz J: Association of cigarettes smoked daily with obesity in a general adult population. Obesity (Silver Spring, Md) 2007, 15(5):1311-1318.
25. Clair C, Chiolero A, Faeh D, Cornuz J, Marques-Vidal P, Paccaud F, Mooser V, Waeber G, Vollenweider P: Dose-dependent positive association between cigarette smoking, abdominal obesity and body fat: cross-sectional data from a population-based survey. BMC public health 2011, 11:23.
8
26. Pasquali R, Vicennati V: Activity of the hypothalamic-pituitary-adrenal axis in different obesity phenotypes. International journal of obesity and related metabolic disorders : journal of the International Association for the Study of Obesity 2000, 24 Suppl 2:S47-49.
27. Chiolero A, Wietlisbach V, Ruffieux C, Paccaud F, Cornuz J: Clustering of risk behaviors with cigarette consump-tion: A population-based survey. Preventive medicine 2006, 42(5):348-353.
28. Baliunas DO, Taylor BJ, Irving H, Roerecke M, Patra J, Mohapatra S, Rehm J: Alcohol as a risk factor for type 2 diabetes: A systematic review and meta-analysis. Diabetes care 2009, 32(11):2123-2132.
29. Ronksley PE, Brien SE, Turner BJ, Mukamal KJ, Ghali WA: Association of alcohol consumption with selected cardiovascular disease outcomes: a systematic review and meta-analysis. BMJ (Clinical research ed) 2011, 342:d671.
30. Schuit AJ, van Loon AJ, Tijhuis M, Ocke M: Clustering of lifestyle risk factors in a general adult population. Preventive medicine 2002, 35(3):219-224.
31. Barefoot JC, Gronbaek M, Feaganes JR, McPherson RS, Williams RB, Siegler IC: Alcoholic beverage preference, diet, and health habits in the UNC Alumni Heart Study. The American journal of clinical nutrition 2002, 76(2):466-472.
32. Sluik D, van Lee L, Geelen A, Feskens EJ: Alcoholic beverage preference and diet in a representative Dutch population: the Dutch national food consumption survey 2007-2010. European journal of clinical nutrition 2014, 68(3):287-294.
33. Ellison RC, Zhang Y, Qureshi MM, Knox S, Arnett DK, Province MA: Lifestyle determinants of high-density lipoprotein cholesterol: the National Heart, Lung, and Blood Institute Family Heart Study. American heart journal 2004, 147(3):529-535.
34. Schrieks IC, Heil AL, Hendriks HF, Mukamal KJ, Beulens JW: The effect of alcohol consumption on insulin sensitivity and glycemic status: a systematic review and meta-analysis of intervention studies. Diabetes care 2015, 38(4):723-732.
35. Castilla P, Echarri R, Davalos A, Cerrato F, Ortega H, Teruel JL, Lucas MF, Gomez-Coronado D, Ortuno J, Lasuncion MA: Concentrated red grape juice exerts
an-tioxidant, hypolipidemic, and antiinflammatory effects in both hemodialysis patients and healthy subjects. The American journal of clinical nutrition 2006, 84(1):252-262.
36. Zern TL, Wood RJ, Greene C, West KL, Liu Y, Aggarwal D, Shachter NS, Fernandez ML: Grape polyphenols exert a cardioprotective effect in pre- and postmenopausal women by lowering plasma lipids and reducing oxida-tive stress. The Journal of nutrition 2005, 135(8):1911-1917.
37. Ebbeling CB, Swain JF, Feldman HA, Wong WW, Hachey DL, Garcia-Lago E, Ludwig DS: Effects of dietary composition on energy expenditure during weight-loss maintenance. JAMA 2012, 307(24):2627-2634. 38. Kraschnewski JL, Boan J, Esposito J, Sherwood NE,
Lehman EB, Kephart DK, Sciamanna CN: Long-term weight loss maintenance in the United States. Inter-national journal of obesity (2005) 2010, 34(11):1644-1654.
39. Bell LK, Edwards S, Grieger JA: The Relationship between Dietary Patterns and Metabolic Health in a Representative Sample of Adult Australians. Nutrients 2015, 7(8):6491-6505.
40. Ortega FB, Cadenas-Sanchez C, Sui X, Blair SN, Lavie CJ: Role of Fitness in the Metabolically Healthy but Obese Phenotype: A Review and Update. Progress in cardiovascular diseases 2015, 58(1):76-86.
41. van Steenkiste B, van der Weijden T, Timmermans D, Vaes J, Stoffers J, Grol R: Patients’ ideas, fears and expectations of their coronary risk: barriers for primary prevention. Patient education and counseling 2004, 55(2):301-307.
42. Weijden T, Timmermans D, Wensing M: “Dus alles is goed dokter” Hoe informeer ik mijn patiënten over grote en kleine risico’s. 2006.
43. Lechner L, Mesters I, Bolman C: Gezondheidspsycholo-gie bij patiënten. Assen: Koninklijke Van Gorcum; 2010. 44. VanderZee KI, Sanderman R, Heyink JW, de Haes H: Psy-chometric qualities of the RAND 36-Item Health Survey 1.0: a multidimensional measure of general health status. International journal of behavioral medicine 1996, 3(2):104-122.
45. Fontaine KR, Barofsky I: Obesity and health-related quality of life. Obesity reviews : an official journal of the
International Association for the Study of Obesity 2001, 2(3):173-182.
46. Heo M, Allison DB, Faith MS, Zhu S, Fontaine KR: Obesity and quality of life: mediating effects of pain and comor-bidities. Obesity research 2003, 11(2):209-216. 47. Katz DA, McHorney CA, Atkinson RL: Impact of obesity
on health-related quality of life in patients with chronic illness. Journal of general internal medicine 2000, 15(11):789-796.
48. Kolotkin RL, Crosby RD, Williams GR: Health-related quality of life varies among obese subgroups. Obesity research 2002, 10(8):748-756.
49. Larsson U, Karlsson J, Sullivan M: Impact of overweight and obesity on health-related quality of life--a Swedish population study. International journal of obesity and related metabolic disorders : journal of the International Association for the Study of Obesity 2002, 26(3):417-424.
50. Soltoft F, Hammer M, Kragh N: The association of body mass index and health-related quality of life in the general population: data from the 2003 Health Survey of England. Quality of life research : an international journal of quality of life aspects of treatment, care and rehabilitation 2009, 18(10):1293-1299.
51. Choo J, Jeon S, Lee J: Gender differences in health-relat-ed quality of life associathealth-relat-ed with abdominal obesity in a Korean population. BMJ open 2014, 4(1):e003954. 52. Fabricatore A, Wadden T: Psychological functioning of
obese individuals. Diabetes Spectrum 2003, 16(4):245-252.
53. Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin BA, Gordon DJ, Krauss RM, Savage PJ, Smith SC, Jr. et al: Diagnosis and management of the metabolic syndrome: an American Heart Association/ National Heart, Lung, and Blood Institute Scientific Statement. Circulation 2005, 112(17):2735-2752. 54. Expert Panel on Detection Evaluation, and Treatment of
High Blood Cholesterol in Adults: Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III). JAMA 2001, 285(19):2486-2497.
55. WHO: Definition and diagnosis of diabetes mellitus and intermediate hyperglycemia: report of a WHO/IDF consultation. In .: World Health Organisation; 2006.
56. Dekker JM, Balkau B: Counterpoint: impaired fasting glucose: The case against the new American Diabetes Association guidelines. Diabetes care 2006, 29(5):1173-1175.
57. Forouhi NG, Balkau B, Borch-Johnsen K, Dekker J, Glumer C, Qiao Q, Spijkerman A, Stolk R, Tabac A, Ware-ham NJ: The threshold for diagnosing impaired fasting glucose: a position statement by the European Diabetes Epidemiology Group. Diabetologia 2006, 49(5):822-827.
58. Flegal KM, Shepherd JA, Looker AC, Graubard BI, Borrud LG, Ogden CL, Harris TB, Everhart JE, Schenker N: Com-parisons of percentage body fat, body mass index, waist circumference, and waist-stature ratio in adults. The American journal of clinical nutrition 2009, 89(2):500-508.
59. Ford ES, Giles WH, Dietz WH: Prevalence of the metabolic syndrome among US adults: findings from the third National Health and Nutrition Examination Survey. JAMA 2002, 287(3):356-359.
60. Kuk JL, Ardern CI: Age and sex differences in the cluster-ing of metabolic syndrome factors: association with mortality risk. Diabetes care 2010, 33(11):2457-2461. 61. Scuteri A, Laurent S, Cucca F, Cockcroft J, Cunha PG,
Manas LR, Raso FU, Muiesan ML, Ryliskyte L, Rietzschel E et al: Metabolic syndrome across Europe: different clusters of risk factors. European journal of preventive cardiology 2015, 22(4):486-491.
62. Alberti KG, Eckel RH, Grundy SM, Zimmet PZ, Clee-man JI, Donato KA, Fruchart JC, James WP, Loria CM, Smith SC, Jr.: Harmonizing the metabolic syndrome: a joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation 2009, 120(16):1640-1645.
63. Han TS, van Leer EM, Seidell JC, Lean ME: Waist circum-ference action levels in the identification of cardiovas-cular risk factors: prevalence study in a random sample. BMJ (Clinical research ed) 1995, 311(7017):1401-1405. 64. Lean ME, Han TS, Morrison CE: Waist circumference as a
measure for indicating need for weight management. BMJ (Clinical research ed) 1995, 311(6998):158-161.
8
65. Florath I, Brandt S, Weck MN, Moss A, Gottmann P, Rothenbacher D, Wabitsch M, Brenner H: Evidence of inappropriate cardiovascular risk assessment in middle-age women based on recommended cut-points for waist circumference. Nutrition, metabolism, and car-diovascular diseases : NMCD 2014, 24(10):1112-1119. 66. Chobanian AV, Bakris GL, Black HR, Cushman WC, Green
LA, Izzo JL, Jr., Jones DW, Materson BJ, Oparil S, Wright JT, Jr. et al: The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. JAMA 2003, 289(19):2560-2572.
67. Zanchetti A, Grassi G, Mancia G: When should antihypertensive drug treatment be initiated and to what levels should systolic blood pressure be lowered? A critical reappraisal. Journal of hypertension 2009, 27(5):923-934.
68. James PA, Oparil S, Carter BL, Cushman WC, Dennison-Himmelfarb C, Handler J, Lackland DT, LeFevre ML, MacKenzie TD, Ogedegbe O et al: 2014 evidence-based guideline for the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA 2014, 311(5):507-520.
69. Mottillo S, Filion KB, Genest J, Joseph L, Pilote L, Poirier P, Rinfret S, Schiffrin EL, Eisenberg MJ: The metabolic syndrome and cardiovascular risk a systematic review and meta-analysis. Journal of the American College of Cardiology 2010, 56(14):1113-1132.
70. Ford ES, Li C, Sattar N: Metabolic syndrome and incident diabetes: current state of the evidence. Diabetes care 2008, 31(9):1898-1904.
71. Reaven GM: The metabolic syndrome: requiescat in pace. Clinical chemistry 2005, 51(6):931-938. 72. Alberti KG, Zimmet P, Shaw J: Metabolic syndrome--a
new world-wide definition. A Consensus Statement from the International Diabetes Federation. Diabetic medicine : a journal of the British Diabetic Association 2006, 23(5):469-480.
73. Eckel RH, Cornier MA: Update on the NCEP ATP-III emerging cardiometabolic risk factors. BMC medicine 2014, 12:115.
74. Wijndaele K, Beunen G, Duvigneaud N, Matton L, Duquet W, Thomis M, Lefevre J, Philippaerts RM: A continuous metabolic syndrome risk score:
util-ity for epidemiological analyses. Diabetes care 2006, 29(10):2329.
75. Smulders YM, Burgers JS, Scheltens T, van Hout BA, Wiersma T, Simoons ML: Clinical practice guideline for cardiovascular risk management in the Netherlands. The Netherlands journal of medicine 2008, 66(4):169-174.
76. Joshi PH, Martin SS, Blumenthal RS: The remnants of residual risk. Journal of the American College of Cardiol-ogy 2015, 65(21):2276-2278.
77. Bell JA, Kivimaki M, Hamer M: Metabolically healthy obesity and risk of incident type 2 diabetes: a meta-analysis of prospective cohort studies. Obesity reviews : an official journal of the International Association for the Study of Obesity 2014, 15(6):504-515.
78. Kramer CK, Zinman B, Retnakaran R: Are metabolically healthy overweight and obesity benign conditions?: A systematic review and meta-analysis. Annals of internal medicine 2013, 159(11):758-769.
79. Achilike I, Hazuda HP, Fowler SP, Aung K, Lorenzo C: Predicting the development of the metabolically healthy obese phenotype. International journal of obesity (2005) 2015, 39(2):228-234.
80. Eshtiaghi R, Keihani S, Hosseinpanah F, Barzin M, Azizi F: Natural course of metabolically healthy abdominal obese adults after 10 years of follow-up: the Tehran Lipid and Glucose Study. International journal of obesity (2005) 2015, 39(3):514-519.
81. Hamer M, Bell JA, Sabia S, Batty GD, Kivimaki M: Stabil-ity of metabolically healthy obesStabil-ity over 8 years: the English Longitudinal Study of Ageing. European journal of endocrinology / European Federation of Endocrine Societies 2015, 173(5):703-708.
82. Appleton SL, Seaborn CJ, Visvanathan R, Hill CL, Gill TK, Taylor AW, Adams RJ: Diabetes and cardiovascular disease outcomes in the metabolically healthy obese phenotype: a cohort study. Diabetes care 2013, 36(8):2388-2394.
83. Phillips CM: Metabolically Healthy Obesity: Personalised and Public Health Implications. Trends in endocrinology and metabolism: TEM 2016, 27(4):189-191. 84. Mann CJ: Observational research methods. Research
design II: cohort, cross sectional, and case-control stud-ies. Emergency medicine journal : EMJ 2003, 20(1):54-60.
85. Scholtens S, Smidt N, Swertz MA, Bakker SJ, Dotinga A, Vonk JM, van Dijk F, van Zon SK, Wijmenga C, Wolffen-buttel BH et al: Cohort Profile: LifeLines, a three-generation cohort study and biobank. International journal of epidemiology 2015, 44(4):1172-1180. 86. Bouter LM, van Dongen MCJM, Zielhuis GA:
Epidemi-ologisch onderzoek: opzet en interpretatie, vol. 5: Bohn Stafleu van Loghum; 2008.
87. Klijs B, Scholtens S, Mandemakers JJ, Snieder H, Stolk RP, Smidt N: Representativeness of the LifeLines Cohort Study. PloS one 2015, 10(9):e0137203.
88. Orho-Melander M: The metabolic syndrome: genetics, lifestyle and ethnicity. Diabetes Voice 2006, 51. 89. Song Q, Wang SS, Zafari AM: Genetics of the metabolic
syndrome. Hospital Physician 2006, 52:51-61. 90. Hassan E: Recall bias can be a threat to retrospective
and prospective research designs. The internet Journal of Epidemiology 2005, 3(2).
