INTRODUCTION
Identification of risk factors contributing to cardiovascular events that are responsible for main morbidity and mortality in our country have always been a real concern for the medical community. Reviewing Jean Vague's works(1) in 1988, Reaven(2) proposed to individualize under the "syndrome X" term the association of arterial hypertension, android obesity, carbohydrates intolerance, dyslipidemia, and hyperinsulinemia reflecting insulin resistance. This entity, named metabolic syndrome, was recently associated with new definitions and growing interest, even considered as a disease in the United States.3
Starting with android obesity, metabolic syndrome seems to be at the intersection of risk factors and an essential actor for cardiovascular events. Very few French studies are available to determine the applicability of these parameters and to assess their clinical outcomes.4,5 Hence, although metabolic syndrome is at the center of numerous meetings, clinical and epidemiological data are dramatically lacking in our country. The Epidemiological Study of Metabolic Syndrome Risk Factors in Military Environment (EPIMIL) offers to fill this gap while demonstrating the feasibility of that kind of survey in the military environment.
METHODS
This 11-year long survey has two goals: to assess frequency and severity of the different cardiovascular risk factors and to gather them within the metabolic syndrome during 1 year and to follow the population concerned during a 10-year period to compare results obtained with forecasts issued from cardiovascular risk calculation and to assess the impact of education measures.
Cross-Sectional Study
Military collectivity provides an ideal environment for this kind of survey because its members all undergo annual routine medical checkups that constitute the right occasion to detect different risk factors. The gendarmerie and Republican Guard from Ile-de-France (near Paris) were already subjected to a transversal survey in 1993, showing feasibility of this type of work owing to motivation and professionalism of doctors from medical units.6 Military formations selected were Republican Guard, antiriot police, and gendarmerie of Ile-de-France because the turnover of personnel is low.
Two thousand subjects were necessary to obtain exploitable statistically significant results. This study was approved by the Begin Hospital Ethics Committee and has been registered with the National Commission for Computers and Freedom. The subjects were informed orally and by letter of the importance of the study and all gave their consent. During an annual examination in 2003, the following information was collected on standard forms.
1. Questionnaire data: biography, toxic habits (tobacco and alcohol consumption), sporting activities, past history and family history. Report of number of cigarettes smoked in pack per years was used to define non smokers, former smokers and active smokers.
2. Alcohol consumption was determined in grams per day from the interview. This approximation proved to be in the past in good correlation with other biological markers of alcohol intake.
3. Examination data: weight and height (for body mass index (BMI) calculation) and waist and hip circumferences measurement. Mean systolic and diastolic blood pressure from two blood pressure readings taken at 5-minute intervals were measured by automatic device on subjects who had been sitting quietly for 5 minutes.
4. ECGs were interpreted by medical staff of the unit, recording any abnormality or left ventricular hypertrophy.
5. Blood and urine samples were taken and sent within 4 hours to the chemistry laboratory of Begin military Instruction Hospital (HIA Begin) for assays. Plasma triglycerides and glucose were measured enzymatically (Olympus reagents, Rungis, France), as was total cholesterol (Biogene reagent, Quimper, France). High-density lipoprotein (HDL) cholesterol was determined by a direct method using polyanions and detergent (Biogene reagent), nonesterified fatty acids were measured by a colorimetric method (Randox reagent, Mauguio, France). All of these parameters were determined by an AU600 analyzer (Olympus). Plasma insulin was measured by Immulite 2000 (Siemens, Puteaux, France). Low-density lipoprotein (LDL) cholesterol was calculated using the formula from Friedewald. The degree of insulin resistance was evaluated by the homeostasis model assessment (HOMA)-fasting insulin [mIU/L] × fasting blood glucose [mmol/L/22.5] and insulin secretion by the HOMA ß cell index (fasting insulin (mIU/L) × 20/ fasting blood glucose (mmol/L) - 3.5).7 Microalbuminuria (Olympus) was assayed on a urine sample. All samples with values >20 mg/L or with a microalbuminuria:creatinine ratio >30 mg/g were considered abnormal. High values of ultrasensitive C-reative protein (CRP; Olympus) were not taken into account.
6. Blood serum and DNA banks were constituted for subsequent assays and genotyping, respectively.
All clinical and biological data were centralized and captured in the Endocrinology Unit of Begin Hospital for their statistical treatment. Criteria defining metabolic syndrome according to the National Cholesterol Education Program (NCEP) Adult Treatment Panel III (ATP III) allowed us to select a group of patients and to compare them to other subjects8 (Table I). Subjects having three or more of the NCEP ATP III criteria were considered affected by metabolic syndrome. Subjects having five criteria were included in the ?4 criteria group. The overall population was then divided into five subgroups. All data are given as means ± SD for continuous variables and numbers of subjects (%) for categorical variables. Dichotomous variables were compared by ?^sup 2^ tests. Student's t test and analysis of covariance were used to compare the other variables. Linear regression was used to determine the correlation between certain parameters. All analyses were done using StatView, version 5.0 (Apple, les Ulis, France). An official protocol to the hospital ethics committee and registration to the National Commission for Computers and Freedom preceded project presentation to the regional direction and to the central Direction of Army health service. Lastly, a meeting for protocol finalization was organized in HIA Begin at the end of January 2003 in the presence of the general, director of Army health service in the Ile-de-France region, authorizing the beginning of the survey for February 1, 2003. Cost of such a study is solely represented by supplementary tests in addition to annual examinations that are left to the judgment of the unit doctor. A financing request on the Army health service research budget was approved for this project.
Follow-Up Study
The follow-up study consists annually of 10 years of both cardiac events and breakdown of risk factors like arterial hypertension, diabetes mellitus, and dyslipidemia. This part of the survey, although clearly difficult to obtain, remains the most informative and requires a plan of contribution of different partners. In case of transfer, a quite rare circumstance in this population, an information letter will be inserted in the medical file, giving instructions to the new attending physician. At retirement time, the patient's whereabouts are noted by a medical secretarial office and a survey is undertaken annually by mail and phone. This procedure should limit loss of patients, all the more because subjects are fully informed of this follow-up study.
Correction measures regarding nicotine addiction, obesity, and dyslipidemia will be set up by unit physicians at the annual examination. Instruction about these risks will be accompanied by dietary intervention. The benefit of these interventions will be assessed by comparison to subjects with no recommendations. This dietician will also be in charge of the patients file management, an essential part of the program. For this portion, Merck, Sharp & Dohme Chibret Laboratories granted funds for 2 years.
RESULTS
Military men (2,045) ages 20 to 58 years old (mean age, 38.6 ± 8.8 years) were recruited. One hundred eight-five (9%) presented with three or more of the five criteria according to the NCEP definition of metabolic syndrome. The differences observed in the parameters were highly significant (Table II), even after correction for age. Elevated blood pressure was the most frequent encountered criteria (51%), before triglycerides (17%), enlarged waist circumference (17%), low HDL cholesterol (9.6%), and high blood glucose (5%). These results did not match exactly to the World Health Organization (WHO) definition.9 Actually, 59 of the 185 subjects described above would be excluded according to the WHO definition, while 162 men that have none of the NCEP definition would be considered having metabolic syndrome according to the WHO criteria. This latter definition is stricter, demonstrating a frequency of 14% for metabolic syndrome for this population instead of 9%. Only 36% are not included by any of the NCEP criteria for metabolic syndrome. Among subjects with the metabolic syndrome, 9.7% had been treated for hypertension, 8.6% had been treated for dyslipidemia, and 3.2% had been treated for diabetes. The comparable figures for metabolic syndrome-free subjects were 1.5%, 4%, and 0.5%.
Patients suffering from metabolic syndrome are significantly older and frequency increased with age. Geographic origin also influenced these percentages, with subjects from Brittany less frequently affected (5%) than those from eastern France (21%).
Patients with NCEP-defined metabolic syndrome had higher BMI and weighed more when they were 20 years old. Smoking, low physical activity, and past history of diabetes or hypertension were more frequently observed in these patients (Table III). Total and LDL cholesterol, microalbuminuria, high-sensitive CRP, and free fatty acids, unlike lipoprotein(a) and homocysteine were higher in patients with metabolic syndrome (Table IV). The plasma insulin and HOMA index that define the insulin resistance level increased regularly and significantly with the number of criteria of metabolic syndrome (Fig. 1). Plasma insulin increased from 6.6 ± 4.2 mIU/L in the absence of criteria to 20.6 ±13 mIU/L when 4 criteria were met (Fig. 1). Insulinemia was very well correlated with BMI (r = 0.456; p < 0.0001) and with waist circumference (r = 0.446; p < 0.0001; Fig. 2). Lastly, several equations for calculating cardiovascular risk showed regular progression of increasing risk as the parameters of metabolic syndrome are met (Table V).10-12
DISCUSSION
These first results on a large cohort confirm the real frequency of metabolic syndrome in a homogeneous population of men. This frequency is lower than that one observed in France and, particularly, in the United States, where it can reach 50% in certain brackets of the population. However, large variations are observed in Europe, depending on criteria that are used or on the population studied.13 Low representation of metabolic syndrome in EPIMIL may be explained by recruitment mode that includes selected and quite young soldiers. Lastly, some subjects particularly at risk refused to participate. This probably explains why that growing frequency of metabolic syndrome with age that is usually observed is not seen here. A lot of these young individuals already show some parameters that define metabolic syndrome. If NCEP and WHO definitions are grouped together, they are somewhat superimposed. Whereas WHO criteria are more heavily weighted with glycemia and insulin resistance, NCEP concentrates on risk factors. Nonetheless, NCEP criteria seem more reasonable because of their simplicity and ease of use.
These results show that metabolic syndrome and insulin resistance are clearly connected. In fact, insulin increases with the appearance of metabolic syndrome criteria and close correlations exist between the insulin level, BMI, blood pressure, and waist size. These observations show that there is no frontier between normality and pathology. Despite imperfections, these observations apply to vascular risk whatever the formula used Actually, findings from the Framingham study in the American population do not take either diabetes or addiction to smoking into account as a risk factor.10 In European subjects, SCORE predicts only mortality and not cardiovascular events.11 In this way, the low rates of young people with important risk factors seems falsely reassuring. Cardiovascular complications are related with metabolic syndrome with increasing highly sensitive CRP or microalbuminuria.14,15 These estimations are strictly correlated with epidemiological observations since having a metabolic syndrome according to WHO criteria in a north European population showed a 3-fold increase in the risk of coronary accident and a 1.8-fold increase in the risk of a fatal cardiovascular event.16
Epidemic obesity clearly effects the incidence of metabolic syndrome, and it may explain why the Bretons, who are less obese than those in the rest of the country are less affected by metabolic syndrome. It precedes and comes with hypertension and moderate hyperglycemia, which is defined by with glycemia higher than 1 g/L.17 The chain of events in which the role of insulin resistance is not completely identified leads to type 2 diabetes, which is associated with significant cardiovascular complications.18 This sequence of events is unfortunately not avoidable with an early monitoring.19 Interventionist surveys have shown the efficacy of change in lifestyle based on dietary modification and increased physical exercise. Several categories of drugs were demonstrated to be effective when a parameter of metabolic syndrome crosses over the threshold for intervention. Hence, metformin, glitazones, statins, adrenocortical extract inhibitors, or aspirin may be used in targeted populations. This potentially fearsome syndrome may be tackled with a better screening, simple preventive measures, and adapted treatment with an expanding panel of medications. Unfortunately, this apparent simplicity of therapy hides grave difficulties that often go beyond the medical field and constitute a sociological problem rather than a simple medical condition.20,21
Beside validation of different equations for calculating cardiovascular risk and clinical and biological assessment for prediction of cardiovascular events, the EPIMIL survey will observe the impact of intervention measures on metabolic syndrome's complications. Military physicians play a major role in prevention. Continuation of this work is in progress with a large study of gene polymorphisms that are known to be involved in glucose or lipids metabolism.22-40 A choice of polymorphisms (single nucleotide polymorphisms, variable number of tandem repeats, or minisatellites) has been made to explore subjects with metabolic syndrome of our population vs. controls (Table VI). A new method using primer extension to allow us to analyze up to 10 polymorphisms in one run is currently in development in our laboratory. Results of genotyping would be useful to better understand the physiopathology of metabolic syndrome.
CONCLUSIONS
There are several advantages for such an epidemiological survey. At the scientific level, it will permit the determination of the frequency of this syndrome in the French Army and assessment of cardiovascular risk for a uniform population of military personnel. Follow-up will precisely identify the role played by each parameter of the metabolic syndrome. DNA collection will be useful at last for genotyping of subjects for several genes possibly involved in the syndrome.
Participation in the survey should make physicians aware of the need for a better following of military personnel. Setup of intensive correction measures should lead to improvement of patient care. Lastly, this survey constitutes an example of the involvement of different elements of the Army Health Service. Hospital and infirmary physicians, pharmacists, and biochemists; nurses and dieticians have combined their expertise and enthusiasm in the EPIMIL group to manage this cost-effective work. This experience shows that with the support of the health service hierarchy and high-quality laboratories, expansion of clinical research is not utopia but works toward a real collaborative practice between hospital and units.
© 2008 Association of Military Surgeons of the United States Provided by ProQuest LLC. All Rights Reserved.
Copyright 2008 Military Medicine
