Five-year predictive factors of type 2 diabetes in men with impaired fasting glucose.

AIM: The outcome of 743 French men (age 20-60) with impaired fasting glucose (IFG) [blood glucose 6.1-6.9 mmol/l] at T1 was evaluated 5 years later, at T2. METHODS: Personal and family medical history, smoking, nutritional habits, physical activity, blood pressure, body mass index (BMI) and waist girth, fasting biological data were collected at T1 and T2. Predictive factors for developing diabetes were compared between those who returned to normal fasting glucose and those who had diabetes, before and after adjustment for age, BMI, glucose and triglyceride (TG) levels. RESULTS: At T2, 44%, 39%, 17% were classified as normal fasting plasma glucose (FPG), IFG or diabetic, respectively. Odd ratios for diabetes were 4.2 for men with a family history of diabetes (FHD), 3.4 if BMI > or = 25 kg/m(2), 2.9 if waist girth > or = 90 cm, 2.8 if TG > or = 2 mmol/l and 1.9 if no daily dairy products were eaten. Still significant after adjustment for age, BMI, glucose and TG levels were: FHD (P=0.001), no daily dairy products (P=0.001), high alcohol intake (P=0.02) and low physical activity (P = 0.02). CONCLUSION: No daily dairy products, high alcohol intake and low physical activity were independent predictive factors of a 5-year onset of diabetes after adjusting for BMI, FHD, triglyceride and glucose levels at baseline. For a better prevention of diabetes, these findings give clues for behaviour modifications as soon as IFG is detected.

Microalbuminuria and markers of the atherosclerotic process: the D.E. S.I.R. study.

The relationship between microalbuminuria and tissue-type plasminogen activator antigen (tPA-ag) and fibrinogen was evaluated in non-diabetic subjects. Subjects were participants of the D.E.S.I. R. (Data from an Epidemiological Study on the Insulin Resistance syndrome) Study. Analyses were carried out on 2248 women and 2402 men for fibrinogen and on 272 women and 284 men for tPA-ag. Microalbuminuria was defined as urinary albumin concentration greater than 20 mg/l. Men with microalbuminuria had a 6% higher fibrinogen concentration than those without (3.07 g/l (95% confidence interval: 2.99,3.15) vs. 2.89 g/l (2.87,2.91), adjusted for age and smoking). This relationship existed in hypertensive as well as non-hypertensive subjects. The association between microalbuminuria and tPA-ag existed only in hypertensive men, those with microalbuminuria having a 21% higher tPA-ag than those without (4.39 ng/ml (3.70,5.08) vs. 3.63 ng/ml (3.32,3.94), adjusted for age and smoking). Adjustment for other risk markers for cardiovascular disease did not change the results. There was no relationship between microalbuminuria and these haemostatic factors in women. The results of this study suggest that in non-diabetic men, microalbuminuria is associated with fibrinogen, but with tPA-ag only when concomitant with hypertension.

Distribution of fasting serum insulin measured by enzyme immunoassay in an unselected population of 4,032 individuals. Reference values according to age and sex. D.E.S.I.R. Study Group. Données Epidémiologiques sur le Syndrome d'Insulino-Résistance.

As hyperinsulinaemia has been shown to be a risk factor for non-insulin-dependent diabetes, cardiovascular disease and hypertension, the measurement of serum insulin levels may provide an additional early screening test for these diseases. Biological assaying of insulin is now facilitated by the IMx-Abbott kit, an enzyme immunoassay which does not cross-react with proinsulin and thus provides more specific insulin determination than conventional methods. Fasting insulin concentrations were determined in a population of 4,032 men and women 30 to 64 years of age, all volunteers for a medical check-up. Concentrations were slightly higher for men in all age-classes (median values of 5.9 and 5.4 microU/ml respectively for men and women). Although significant differences were found in serum insulin concentrations between the four age-classes for men and women, there were no significant differences between the three age-classes for men up to 59 years (median: 5.8 microU/ml, 95th percentile 14.0 microU/ml) or between the two age-classes for women up to 49 years (5.2, 12.5). Fasting concentrations were increased above these age thresholds: men (6.4, 15.6), women (5.6, 14.0). The reference population consisted of 3,081 non-diabetic, glycosuria-negative subjects with a body mass index and glucose concentration lower than the 95th percentiles for their age and sex. The reference values for fasting insulin concentrations were: 1) women 30-49 years: median 5.1 microU/ml (95% confidence interval: 4.9-5.3), 95th percentile 11.2 microU/ml (10.9-11.9); and 2) men 30-64 years and women 50-64 years: median 5.6 microU/ml (5.4-5.7), 95th percentile 12.6 microU/ml (12.0-13.0).

Impaired fasting glycaemia and undiagnosed diabetes: prevalence, cardiovascular and behavioural risk factors.

BACKGROUND: Early discovery of type 2 DM (NIDDM) is essential. The diagnostic criteria of DM have been recently modified (FBG 126 vs 140 mg/dl) and the characteristics of undiagnosed subjects in large populations must be defined. At the same time subjects with impaired FBG need to be studied mainly for their cardiovascular complications. METHODS: During 14 months, 61,724 male and female subjects (mean age 40) were explored in the French Institute for Health Protection (I.R.S.A). Clinical data, FPG, CV risk factors and dietary habits collected. Cut-off value for FPG: 110-125 mg/dl (IFG) (G2), 126-139 mg/dl defining undiagnosed diabetes with no history of diabetes. Subjects with FPG >=140 mg/dl (G4) former ADA/WHO criteria for diabetes and with the new criteria (FPG: 126-139 mg/dl) (G3) were compared to IFG (G2) and controls<110 mg/dl (G1). RESULTS: With the new criteria (>=126 mg/dl) the prevalence of unknown diabetes in the cohort was 1.2% accounting for 41% of the overall prevalence of the disease (known + unknown). This is nearly 2.5 times more than with the previous criteria, > 140 mg/dl, (1.2 vs 0.5%). In G2/G1 and G3/G2 highest FPG had higher BMI, H/W ratio, heart rate (male only G3/G2), BP, gamma GT (role of alcohol in males), uric acid and TG. A role of absence of breakfast, low dairy products consumption is found. No difference between G4 and G3 found. CONCLUSION: These results support the new criteria of FPG 126 mg/dl and suggest that it would be necessary to investigate and prevent cardiovascular risk factors as soon as fasting glycaemia is found to be over 110 mg/dl. Nutritional and behavioural education should be given at this early stage of the disease.

Alcohol intake and fasting insulin in French men and women. The D.E.S.I.R. Study.

BACKGROUND: To study the relation between alcohol consumption and the fasting insulin concentration in a French population with a range of alcohol intakes. METHODS: 2.406 men and 2.500 women, aged 30 to 65 years who were not known as diabetic and with a non-diabetic fasting plasma glucose<7.0 mmol/l were studied. Insulin was assayed by a specific micro-enzyme immunoassay and alcohol intake was from a self-questionnaire. RESULTS: Fasting insulin concentration showed an inverse linear association with alcohol consumption, after adjustment for age and possible confounding factors (p for trend<0.0001 men; p<0.002 women), with a 29% higher insulin in non-drinkers compared to very heavy drinkers (> 80 g/day) in men (p<0.0001) and a 23% and 26% difference when compared to heavy drinkers (41-80 g/day) in men and women respectively (p<0.0001, p<0.003). This relation did not differ significantly according to whether the alcohol was consumed as wine, beer/cider or spirits. Fasting plasma glucose modified the relation between alcohol and insulin in men: while the negative relation alcohol-insulin was strong for fasting plasma glucose<6.0 mmol/l (p<0.0001), there was no association above 6.0 mmol/l (p=0.4). CONCLUSION: There is an inverse relation between alcohol consumption and fasting insulin concentrations. Some studies have found a U shaped relation, and this is probably due to the inclusion of diabetic subjects. As hyperinsulinemia has been shown to be positively associated with cardiovascular disease, it may be one of the variables that explains the protective effect of moderate alcohol consumption on cardiovascular disease.

Comparison of men with impaired fasting glycaemia to controls and to diabetic subjects with fasting glycaemia from 7.0 to 7.7 mmol/l: clinical, nutritional and biological status.

OBJECTIVE: To compare medical history, clinical, nutritional and biological status of non-diabetic men to subjects with impaired fasting glycemia (glycemia 6.1-6.9 mmol/l) and to newly diagnosed type 2 diabetic subjects (7.0-7.7 mmol/l) according to the criteria proposed by the American Diabetes Association. METHODS: Cross-sectional study of a cohort of 29,992 men, who were volunteers for a free periodic check-up offered by their medical insurance. Medical history, lifestyle and nutritional habits were recorded using a self-administered questionnaire. Clinical and biological data were also studied. To compare the three groups of subjects - normal, impaired fasting glycemia and newly diagnosed diabetics - three age stratified samples were randomly designed. RESULTS: Most of the well-known risk factors for developing type 2 diabetes mellitus such as overweight, abdominal obesity, familial history of diabetes mellitus, over-consumption of fat and alcohol were present in the group with impaired fasting glycaemia which presented the same risk factors as the group of subjects with fasting glycaemia from 7.0 to 7.7 mmol/l, but to a lesser degree. Hypertension was present in more than 50% of the subjects with impaired fasting glycaemia. CONCLUSION: In this cross-sectional study, impaired fasting glycaemia is associated with the risk factors of type 2 diabetes mellitus. The subjects with impaired fasting glycaemia should be considered at risk for cardiovascular disease and might take advantage from early specific intervention about their lifestyle.

Insulin dose and cardiovascular risk factors in type 1 diabetic children and adolescents.

Associations between a high daily insulin dose and cardiovascular risk factors, including those of the insulin-resistance syndrome, were studied in 479 Type 1 diabetic children 6 to 18 years of age. Insulin dose increased over the first two years after diagnosis of diabetes (p = 0.0001) and was significantly higher in girls (p = 0.01). For those children with diabetes duration of more than 2 years, the insulin requirement increased up to 13-14 years of age (p < 0.05) and was higher in pubertal than pre-pubertal children (p < 0.05). For girls, the requirement was higher in puberty than in post-puberty (p < 0.05) and increased with diabetes duration (p < 0.05). Triglyceride concentrations were associated positively and significantly with the insulin dose of both boys and girls, after adjustment for age, pubertal stage, diabetes duration, and metabolic control (fructosamine levels). No other consistent associations were found between insulin dose and other cardiovascular risk factors: body mass index, central adiposity, arterial blood pressures, serum total cholesterol, apoA1, apoB, Lp(a), uric acid, or urinary albumin excretion. Parental obesity, hypertension and diabetes were not related to the insulin dose of children. The results did not differ when the population was limited to the 375 children with diabetes duration of more than 2 years. It is concluded that in these Type 1 diabetic children the insulin dose for a given level of metabolic control (our surrogate measure of insulin resistance) was related to a single cardiovascular risk factor: triglyceride concentrations.

A screening method for abnormally high lipoprotein(a) concentrations by agarose lipoprotein electrophoresis.

A routine electrophoretic method detecting plasma lipoprotein(a) (Lp(a)) is described. Plasma lipoproteins were electrophoresed using an agarose gel film containing cations which retard migration of beta-, prebeta- and alpha-bands. When present, the Lp(a)-band was detected between prebeta- and alpha-bands. This extra-band lipoprotein has been demonstrated to be Lp(a), by an immunofixation technique using anti-Lp(a) antibodies. This original procedure allows a distinct separation of Lp(a) from prebeta even after samples have been stored at 4 degrees C for several days, or in cases of hyperlipemic samples with increased prebeta lipoproteins. The reliability of this detection test has been tested in comparison with an Lp(a) electroimmunoassay. Both these techniques have been performed on 719 randomly selected subjects. With electrophoresis, the Lp(a)-positive subjects accounted for 34.2% of the subjects and although this method does not distinguish between different levels of positivity (depending on the sample), the presence of Lp(a)-band was always perceptible at concentrations that belong to the upper 15th percentile of values as determined by electroimmunodiffusion; inversely, all Lp(a)-positive plasma was measurable. In consequence, since it is reliable and relatively inexpensive, this detection test on modified agarose gel appears very useful for revealing the presence of abnormally high values of Lp(a) in populations.

[Is smoking history a risk factor of arterial hypertension in men?]. / Le tabagisme ancien est-il un facteur de risque d'HTA chez l'homme?

Former smokers exhibit decreased cardiovascular risk as compared to smokers who continue to smoke. However, smoking discontinuation results in weight gain which may be important and influence arterial pressure. From January 1st to June 30th, 1998, 12,417 volunteers (aged 20 to 69) were examined at the "Institut régional pour la santé" (IRSA, Regional Institute for Health), a group of 9 social medical centres in Western and Central France. The subjects were screened for a routine medical and biological check-up provided by their medical insurance. All of the subjects were interviewed by a trained nurse who completed a standardised questionnaire regarding personal medical history, current treatments and lifestyle behaviours (especially alcohol and smoking habits). A physician recorded clinical parameters including age, weight, height, systolic and diastolic arterial pressure. Body mass index (BMI) was calculated. Non smokers and former smokers represented 40.0% and 23.8% of the population respectively. The prevalence of a BMI 27.0 kg/m2 or greater was higher in former smokers than non smokers and current smokers. Systolic and diastolic arterial pressure in former smokers exceeded those of current smokers and non smokers by 4.2/1.1 mmHg and 2.8/1.6 mmHg respectively. Using logistic regression analysis, the relative risk of hypertension in former smokers was 1.24 (CI 95%: 1.10-1.39, p < 0.001) and 1.13 (0.995-1.29, p = 0.055) as compared to non smokers and current smokers, after adjustment for age and alcohol intake. Differences became non significant when BMI was entered in the model. The results of the present study suggest that former smoking status is associated with a higher prevalence of overweight which may cause a higher prevalence of hypertension.

[Seroepidemiology of hepatitis A in six departments in West-Central France in 1991]. / Séroépidémiologie de l'hépatite A dans six départements du Centre-Ouest de la France en 1991.

Antibodies to hepatitis A virus were sought for among residents of 6 "départements" of West-Central France during a routine medical check up provided by the national health insurance system to all affiliated persons. Among them 5,641 subjects (aged 6 to 60) were randomly selected to a sample size of 256 +/- 15 male or female subjects for each five-year age group. Anti-HAV were detected in 51% of the screened population. The prevalence was less than 5% in the 6-15 age group and increased by successive steps to exceed 90% in the subjects over 50. The highest increase was observed between the 26-30 (39%) and the 31-35 (57%) age groups. This study confirms the influence of socio-economical factors on the anti-HAV prevalence rate. Factors such as profession or scholar education were previously known, but the housing conditions and especially the number of siblings were identified in our present report. In relation to this low prevalence of anti-HAV among young adults, one may fear an increase of overt hepatitis cases which are known to be more frequent in adults than in infants or children. Moreover any break in hygiene may expose the population to an epidemia. Only active immunization with a vaccine against hepatitis A could efficiently prevent these individual and community-related risks.

[Reference values of apolipoproteins A1 and B. Contribution of international standardization. Travail collaboratif entre la SFBC, l'Arcol et le SFRL]. / Valeurs de référence des apolipoprotéines A1 et B. Apport de la standardisation internationale. Travail collaboratif entre la SFBC, l'Arcol et le SFRL.

The utilization of two WHO reference materials, liquid and lyophilized, permitted international standardization of apolipoprotein measurements. We report here the results of a collaborative study between Arcol, SFBC and SFRL in order to establish reference ranges for apo A1 and B on nine standardized systems. A population of 1027 men and women supposed healthy, 4 to 60 year old, have been selected in two Centers for Preventive Medicine. The serum samples were aliquoted frozen at -20 degrees C the day of sampling and analysed by the manufacturers with IFCC standardized calibrants. A specific quality control was performed using a frozen pool of sera. For apo A1, the centile 2.5 of the reference population varies from 1.04 to 1.16 g/l. The range values for the centile 97.5 varies from 1.87 to 2.24 g/l. For apoB, the centile 2.5 varies from 0.43 to 0.57 g/l, and the centile 97.5 from 1.30 to 1.39 g/l. Only one system has a problem of dispersion with an upper limit equal to 1.20 g/l. These results improve that international standardization allowed actually a good comparability of the results, especially for apoB.

[Use of 6 serum pools in standardization of the determination of apolipoproteins AI and B in the human serum]. / Utilisation de 6 calibrants sériques pour standardiser les dosages des apolipoprotéines AI et B dans le sérum humain.

The proliferation of commercial immunoassays for the determination of apolipoproteins AI and B as parameters of risk for coronary heart disease and the discordance between the values obtained with the systems have made more urgent than ever the need for standardization of these assays. The authors report here the results of a collaborative study on the standardization of the values between the "Comité Française de Coordination des Recherches sur l'Athérosclérose et le Cholestérol (ARCOL)" and 13 Companies providing 15 different analytical systems. The results show that standardized measurements are not possible using a "consensus international standard" and that the major problem is related to the enormous variability of the analytical systems. In contrast, the present study demonstrates that six different serum pools containing increasing amounts of apolipoproteins AI and B can be used to generate a common reference curve which permits calibration of all immunoassays and standardization of results.

[Reference values of lipoprotein(a) in a French population]. / Valeurs de référence de la lipoprotéine(a) dans une population française.

OBJECTIVES: The aim of this work is to study the effect of different biological factors that could affect Lp(a) level in a presumably healthy population and to establish reference limits. METHODS: We selected 723 subjects (367 men and 356 women) for the age interval 4 to 64 years for evaluation. RESULTS: The distribution of Lp(a) is not Gaussian; 50.5% of subjects had Lp(a) concentrations under 0.10 g/l and the value for the 75th percentile was 0.27 g/l and 0.57 g/l for the 90th percentile. No relationship was observed between Lp(a) concentration and cholesterolaemia, triglyceridaemia, glycaemia, inflammatory proteins (orosomucoide and CRP), overweight, tobacco consumption and oral contraceptive use. The menopause state in women was a factor correlated with increased Lp(a) but this increase was not significant. Moreover, alcohol consumption (more than 44 g per day in men and more than 22 g per day in women) was associated with lower Lp(a) values. Among familial cardiovascular risks, only paternal listing of hypertension was associated with Lp(a) concentration in men. CONCLUSION: The measurement of Lp(a) in a young subject could be used as a genetic marker of cardiovascular risk associated with abnormal lipid metabolism and thrombosis phenomena.

Relations among night work, dietary habits, biological measure, and health status.

A cross-sectional study was conducted to investigate dietary intake, behavioral habits, and clinical and metabolic differences in night workers compared to day workers and to evaluate the metabolic differences associated with diet and body habits that occur between these two groups. Dietary habits, biological parameters, and health status were collected in 1,200 night workers and in an equal number of day workers, matched for gender, age, and socioeconomic status. Our findings demonstrated that night workers had poorer dietary habits and metabolic profile compared to day workers with a similar overall health status. These differences were associated with a higher prevalence of some cardiovascular risk factors such as smoking and obesity.

Reference limits of apolipoprotein A-I and apolipoprotein B using an IFCC standardized immunonephelometric method.

An important collaborative study organized by the IFCC enabled the selection of international reference materials and the standardization of commercially available methods by the use of common calibrators. In this paper, we report the reference limits of apolipoprotein A-I and apolipoprotein B in a selected healthy French population. The apolipoprotein measurements were performed on BNA Behring using reagents and protocols supplied by the manufacturer: the standard sera were calibrated using the IFCC-WHO reference preparations (SP1 and SP3). In addition, the apolipoprotein B protocol was modified by the addition of a supplementary reagent to reduce the interference by lipaemic samples on immunonephelometric measurement. In a sample of 115 random serum samples, there was a decrease in mean concentration between non-standardized and standardized methods: -4.8% for apolipoprotein A-I and -4.7% for apolipoprotein B. The reference limits for apolipoprotein A-I are unaffected by gender between 4 and 14 years, thereafter vary with age and gender until 40 years and with gender alone between 40 and 60 years. For apolipoprotein B, the variation with gender is only significant between 30 and 49 years.

Nephelometric determination of carbohydrate deficient transferrin.

We describe a technique for measuring carbohydrate-deficient transferrin (CDT) in serum. Serum transferrin fractions are separated by anion-exchange chromatography on microcolumns. Sialic acid-deficient transferrin fractions are collected in the eluate, and transferrin is then quantified by a rate-nephelometric technique. Imprecision (CV) was 4-5% within-run and 7-9% between runs (n = 15). Comparison with an isoelectric focusing-immunofixation method for transferrin index (x) yielded y = 761x + 7, Sy/x = 39 mg/L. Assay of sera from 90 abstainers or moderate consumers of alcohol showed that 81 (90%) had CDT concentrations between 30 and 70 mg/L. Among 74 alcoholics admitted to an alcohol treatment center, 54 (73%) had CDT > 70 mg/L, i.e., the diagnostic sensitivity was 73% at a specificity of 90% (area under receiver-operator characteristic curve = 0.891).

Fibrinogen: a possible link between alcohol consumption and cardiovascular disease? DESIR Study Group.

The relation between alcohol consumption and fibrinogen concentration was evaluated in a French population to investigate whether fibrinogen could explain part of the relation between alcohol consumption and cardiovascular disease. Cross-sectional data on self-reported alcohol consumption and fibrinogen, measured by the immunonephelometric method, of 4967 men and women aged 30 to 64 years were used. These subjects were volunteers for a free health checkup in the western central part of France from 1994 to 1996 and participated in the DESIR Study (Data from an Epidemiological Study on the Insulin Resistance syndrome). Alcohol consumption was strongly associated with fibrinogen concentration, with higher concentrations in those who were nondrinkers or who drank >60 g of alcohol per day. This U-shaped association was stronger among men than women. Consumption of wine and spirits was associated with fibrinogen, whereas consumption of beer or cider was not. Furthermore, smoking was positively associated with fibrinogen concentration, and in men the difference between nondrinkers and drinkers with the lowest fibrinogen level was higher in nonsmokers and ex-smokers than in current smokers. We conclude that moderate drinking may lower fibrinogen concentration. If fibrinogen is a causal risk factor for cardiovascular disease, it may be 1 of the variables that explain the protective effect of moderate alcohol consumption on cardiovascular disease.

Hormonal status and NIDDM in the European and Melanesian populations of New Caledonia: a case-control study. The CALedonia DIAbetes Mellitus (CALDIA) Study Group.

OBJECTIVE: To assess ethnic differences in androgenic status related to non insulin-dependent diabetes mellitus (NIDDM) in male and female Melanesians and Europeans of New Caledonia. DESIGN: This is a case-control study nested in a prevalence study for diabetes mellitus in the multiracial population of New Caledonia. SUBJECTS: 186 male subjects were included in the survey (77 Melanesians and 16 Europeans in each case and control group). Each case and control group included 104 female Melanesian subjects (69 premenopausal and 35 postmenopausal). METHODS: Diabetic subjects were matched for age, gender, ethnic group and location, with healthy normoglycaemic subjects. Testosterone levels in men and sex hormone-binding globulin (SHBG) levels in women (measured by radioimmunoassay, RIA) were compared between NIDDM and control subjects in relation to obesity, central adiposity and insulin levels. RESULTS: In both ethnic groups, NIDDM was associated with lower testosterone levels but there was a marked difference among Europeans. Testosterone was negatively associated with the body mass index (BMI) (r= -0.35, P <0.01) and fasting insulin (r= -0.37, P <0.001) in control Melanesians only. In Melanesian women, NIDDM was associated with lower SHBG levels in pre- and postmenopausal women (P <0.001). SHBG mean level was not associated with menopausal status. CONCLUSION: Our results confirm in a Pacific population that NIDDM is associated with low levels of testosterone in men and low levels in SHBG in women. In contrast to white populations, Melanesian women have a more androgenic profile, whatever their menopausal status.