The Role of Genetics in Defining Reference Values and Health Status

Since its establishment, the Center for Preventive Medicine in Vandoeuvre-les-Nancy, France, performed specific studies on healthy humans, and its approach was very useful for defining reference values. Prevention should extend its interest to chronic diseases. The majority of important adult disorders are partially genetically determined. Genetic markers are also useful as exclusion or as partition criteria in the production of reference values. Results are presented that were obtained for apolipoproteins E, B and AIV, frequencies of these polymorphisms in the Lorraine population, and relationships between these polymorphisms and lipid metabolism-related parameters. Health checkup centers, in particular those involved in family screening, are well suited for reassembling many data concerning environmental factors: tobacco consumption, alimentation habits, or alcohol and drug consumption. Simultaneous determination of genetic markers could allow the determination of an individual’s susceptibility or resistance to developing a disease and to prepare a preventive action.

Association of human cathelicidin (hCAP-18/LL-37) gene expression with cardiovascular disease risk factors

BACKGROUND AND AIMS

Antimicrobial peptides (AMPs) are components of the innate immune system. In addition, evidence suggests that these peptides are associated with various inflammatory diseases. We examined whether expression of the cathelicidin LL-37 in peripheral blood mononuclear cells (PBMCs) is associated with cardiovascular risk factors.

METHODS AND RESULTS

A total of 90 men and 87 women selected from STANISLAS cohort were studied. Expression of LL-37 mRNA isolated from PBMCs of these subjects was quantified by quantitative RT-PCR. Anthropometric measurements and biochemical profiles were assessed for each individual. In women, LL-37 mRNA expression was significantly and positively correlated with body mass index (BMI) (p<or=0.001); waist circumference (WC) (p<or=0.01); systolic blood pressure (SBP) (p<or=0.05) and triglycerides (TG) level (p<or=0.05) and negatively with plasma levels of HDL-C (p<or=0.05). In men however, LL-37 was positively associated with waist to hip ratio (WHR) (p<or=0.05); SBP (p<or=0.001); TG (p<or=0.05); fasting glucose levels (p<or=0.01); alanine aminotransferase (ALT) activity (p<or=0.01); neutrophils counts (p<or=0.01) and negatively with lymphocyte counts (p<or=0.001); serum HDL-C (p<or=0.001) and apoA-I (p<or=0.05) levels. After adjustment for WC and BMI, multiple regression analysis showed that LL-37 remained significantly associated with SBP; HDL-C; fasting glucose level; ALT activity; neutrophil and lymphocyte counts (p<or=0.001 to p<or=0.05) in men.

CONCLUSION

Our results suggest that LL-37 gene expression may be closely associated with cardiovascular risk factors independently of BMI and WC. However, functional studies are required to confirm these data.

P-selectin polymorphisms' influences on P-selectin serum concentrations and on their familial correlation: the STANISLAS family study

BACKGROUND

P-selectin is an adhesion molecule known to be involved in the pathogenesis of several diseases through its major role in the initial phase of leukocytes recruitment during inflammation. However, genetic characterization of soluble P-selectin remains unclear.

OBJECTIVES

In the STANISLAS cohort, we study the familial correlations of P-selectin levels and investigate the association of six P-selectin polymorphisms (C-2123G, A-1969G, S290N, N562D, V599L and T715P) and cardiovascular risk factors with P-selectin concentrations.

PATIENTS/METHODS

Full phenotypic and genotypic information was available for 136 healthy families composed of both natural parents and at least one child (boys, n = 125; and girls, n = 139) aged more than 4 years.

RESULTS

While no correlation was observed between spouses, family correlations of P-selectin concentrations were highly significant for sibling (0.50 +/- 0.12, P < 10(-3)) and child-parent pairs (0.42 +/- 0.04, P < 10(-3)). P-selectin haplotypes explained about 25% of the variability of P-selectin concentrations, this effect being mainly due to the additive effects of two polymorphisms, V599L and T715P. After adjusting for the effect of the P-selectin polymorphisms, the sibling and child-parent correlations decreased to (0.39 +/- 0.08, P < 10(-4)) and (0.32 +/- 0.06, P < 10(-4)), respectively.

CONCLUSIONS

In the present study, we showed that two P-selectin polymorphisms, V599L and T715P, explained about 25% of the variability of P-selectin concentrations and accounted for about 40% of their family resemblance. These results would suggest a genetic influence on P-selectin concentrations beyond the contribution of the P-selectin gene.

Left ventricular systolic dysfunction is an independent predictor of homocysteine in angiographically documented patients with or without coronary artery lesions

BACKGROUND

Elevated plasma homocysteine is a risk factor for coronary artery disease (CAD) and thromboembolic disorders that seems also to be associated with chronic heart failure.

OBJECTIVE

To evaluate the association between homocysteine and left ventricular dysfunction and to assess whether it is independent of CAD.

PATIENTS AND METHODS

A prospective study evaluated this relationship in 709 patients referred for diagnostic coronary angiography, including 515 CAD and 194 patients without evidence of coronary artery lesions.

RESULTS

The homocysteine level was significantly higher in the 187 patients with a left ventricular ejection fraction (LVEF) dysfunction < 40% (P < 0.0001) than in those without ventricular dysfunction. LVEF, NYHA functional class II or III and CAD, stable angina and hypertension were clinical characteristics that influenced total homocysteine level in univariate analysis. Homocysteine was significantly associated with LVEF and N-terminal pro-B-type natriuretic peptide (NT-pro-BNP) in univariate regression (r = -0.267, 95% CI -0.33 to -0.19, P < 0.0001, and r = 0.381, 95% CI 0.28-0.47, P < 0.0001, respectively) and in multiple regression (P = 0.0022 and P = 0.0001, respectively). Other determinants were creatinine and vitamin B(12), but not folate. LVEF was a predictor of homocysteine > 15 micromol L(-1) in the whole population (P for trend < or = 0.0001) and in patients without documented CAD (P for trend = 0.0058).

CONCLUSION

Our results showed an association of homocysteine with left ventricular systolic dysfunction and NT-pro-BNP that existed independently of documented CAD. Whether this association reflects a causative factor or a consequence of CHF and influences the prognosis of the disease remains an open question.

Heritability of serum hs-CRP concentration and 5-year changes in the Stanislas family study: association with apolipoprotein E alleles

We aimed at estimating additive genetic heritability, household component effect and the influence of common alleles of the apolipoprotein E gene (APOE) on serum high-sensitivity C-reactive protein (hs-CRP) concentrations and the subsequent changes over 5 years. A sub-sample of 320 nuclear families was randomly selected from the Stanislas Family Study. Serum hs-CRP concentration was measured by immunonephelometry at entrance and after 5 years. APOE alleles were determined by restriction fragment length polymorphism. After adjustment for covariates, the number of the epsilon4 allele was negatively associated with serum concentration of hs-CRP in the whole sample, at entrance and 5 years later, without significant interaction with sex by generation groups (P=0.003 and P=0.0003, respectively). However, no significant association was found between epsilon4 allele and 5-year changes in hs-CRP concentration. Using a variance component analysis, no significant genetic influence was shown in family aggregation of both hs-CRP measurements and 5-year changes; the household common component was between 6.5 and 12.8%. In addition, after adjustment for APOE gene polymorphisms, degrees of resemblance were almost unchanged. In the Stanislas Family Study, epsilon4 allele of the APOE gene was associated with lower hs-CRP concentration, but not with 5-year changes. However, variance component analysis did not evidence a significant polygenic effect.

[Glutathione S-transferases genetic polymorphisms and human diseases: overview of epidemiological studies]

Glutathione S-transferases (GST), xenobiotic-metabolising enzymes, are involved in the metabolic detoxification of various environmental carcinogens. Particular genetic polymorphisms of these enzymes have been shown to influence individual susceptibility against various pathologies including cancer, cardiovascular and respiratory diseases. The results from the meta-analysis indicate that GSTM1*0 null allele was associated with enhanced risk for lung (OR (95% IC) = 1,17 (1,07-1,27)), bladder (OR = 1,44 (1,23-1,68) and larynx cancer (OR = 1,42 (1,10-1,84)). GSTT1 null genotype was associated with increased astrocytomas (OR = 2,36 (1,41-3,94)) and meningiomas (OR = 3,57 (1,82-6,92)) cancer risk. GSTP1 allelic polymorphism influence the development of bladder cancer in smokers (OR = 2,40 (1,12-4,95)) and occupational asthma (OR = 3,5 (2,7-4,6)). Finally, GSTM1*0 null allele and GSTT1*1 functional allele were associated with increased risk for coronary heart diseases in smokers (OR = 2,30 (1,40-9,00)) and OR = 2,5 (1,30-4,80), respectively). The GSTT1*1 functional allele was also significantly associated with increased risk of lower extremity arterial disease (OR = 3,60 (1,40-9,00). These epidemiological data suggest that genetic GST polymorphisms influence the individual susceptibility to these diseases. Contrary to cardiovascular disease, no evidence of interaction between GST genotype and smoking status was found in lung cancer but it has not been studied in other cancers. Consequently, other works are necessary to study the potential interaction between GST genotype and environmental carcinogens including tobacco smoke extract.

Family resemblance in breakfast energy intake: the Stanislas Family Study

BACKGROUND

There seems to be a consensus that family influences on dietary habits are important. However, no data relative to breakfast have been published yet.

OBJECTIVE

To investigate whether and how breakfast energy intake aggregates within French families.

DESIGN

A total of 398 families of the Stanislas Family Study who filled in a 3 day food consumption diary were selected. Absolute and relative breakfast energy intakes (BEI in kcal/day and RBEI in percentage of daily intake, respectively) were both studied.

RESULTS

By using a variance component analysis, no genetic influence was shown in family aggregation of both BEI and RBEI. Intra-generation common environmental contribution to total phenotypic variance of BEI and RBEI was higher than inter-generation; both were increased with frequency of sharing breakfast. Furthermore frequency of sharing breakfast contributed to increase family resemblance in breakfast energy intake, particularly in offspring for BEI and RBEI, and in spouses for RBEI. Smoking habits, alcohol consumption, BMI or physical activity were related to family resemblance, but after adjustment on each factor degrees of resemblance were almost unchanged.

CONCLUSION

General findings of this study were that family aggregation in breakfast absolute and relative energy intakes was significant within Stanislas families. Family resemblance depended on inter- and intra-generation components and was modified by the number of shared breakfasts. Our study confirmed that familial habits act on family resemblance in both absolute and relative breakfast energy intakes, so that family should be a favorite unit for health and diet promotion programs.

SPONSORSHIP

Kellogg's PA, France.

High sensitivity C-reactive protein (CRP) reference intervals in the elderly

High sensitivity C-reactive protein is a useful marker in clinical practice; however, reference intervals are not available for all age groups. Therefore, the aim of this study was to determine reference intervals for elderly people. The non-parametric reference limits were calculated for the two genders, subdivided into two age classes (50-64 and 65-91 years). In our selected sample population, we did not observe significant gender differences.

Soluble transferrin receptor (sTfR): biological variations and reference limits

The aim of this study was to establish soluble serum transferrin receptor (sTfR) reference limits. sTfR was measured in 885 healthy subjects from 3 to 91 years old (433 men, 409 women), without hematological abnormalities, using an immunonephelometric assay. The sTfR median concentrations in our population decreased gradually from the group aged 3-10 years to the group aged 21-40 years, then there were no changes in the older groups except for the females >60 years of age. The interindividual variability ranged from 12.6% to 30.3% among different age groups, and the analytical variability was 5%. Biological factors and other factors associated with sTfR concentration variation were examined and accounted for 35% of the sTfR variability in men aged 20 years or less, and 18% in those older than 20 years. Also, they accounted for 45% of the variability in women aged 20 years or less and 14% in those older than 20 years. The main factors statistically associated with sTfR concentration in males were ferritin, orosomucoid, hemoglobin, and tobacco in all age groups and only mean corpuscular volume (MCV) in males less than 20 years old. In the females the main factors were age, orosomucoid, and hemoglobin in all age groups, MCV and tobacco in females less than 20 years old, and ferritin and physical activity in females more than 20 years old. These factors were used to define the exclusion and partition criteria for obtaining the reference samples. Medians for reference values were: 1.60 mg/l in the 3-10-year old group (males and females); 1.42 mg/l in males between 11 and 20 years of age, and 1.33 mg/l in females of the same age. In the other age groups, the median of the reference values was 1.16 mg/l, except in females over 60 years old, for whom it was 1.26 mg/l.

LEPR gene polymorphisms: associations with overweight, fat mass and response to diet in women

BACKGROUND

In humans, a mutation of the leptin receptor gene (LEPR) leads to a rare obese syndrome of mendelian inheritance. However, obesity in humans results from interactions between genes and environment, mainly nutritional factors. Variations at the LEPR locus could be involved in the regulation of body weight.

DESIGN

Genetic variations at the LEPR locus were screened in a selection of 30 French overweight subjects by Single Strand Conformation Polymorphism (SSCP) analysis, then an association study between genotypes and obesity phenotypes was performed in 179 French overweight patients recruited from the Nutrition Department of Bichat Hospital in Paris who were prescribed a low calorie diet and in 387 unrelated volunteers (98 overweight, 289 normal weight) drawn from the Stanislas Family Study in Nancy.

RESULTS

Two new genetic variants were found: T + 70-->C (exon 1) and Asp (A) 96 Asp (G) (exon 4). In Nancy, the T + 70-->C polymorphism was associated with fat mass adjusted for BMI in women (P = 0.025). The genotype and allele frequencies of the Ser (T) 343 Ser (C) polymorphism (exon 9) were significantly different between normal and overweight women, with the T allele being more frequent in the overweight group (T frequency in Nancy, 0.82; in Nancy + Paris, 0.79) than in the normal weight group (0.69; P = 0.017 vs. Nancy overweight, P = 0.003 vs. Nancy + Paris overweight). In women from Nancy, fat mass adjusted for BMI was significantly associated with this polymorphism (P = 0.01). The overweight women carrying the C allele of this polymorphism lost more weight in response to low calorie diet than the non carriers (P = 0.006).

CONCLUSIONS

In women, genetic variations at the LEPR gene level are associated with overweight and fat mass in a cross sectional study and with response to low calorie diet in an intervention study. These results indicate that variations at the leptin receptor locus are associated with common obesity phenotypes and are a part of the polygenic influences on the response to nutritional environment.

Apolipoproteins E and C-III in apo B- and non-apo B-containing lipoproteins in middle-aged women from the Stanislas cohort: effect of oral contraceptive use and common apolipoprotein E polymorphism

Oral contraceptive (OC) use and common apo E polymorphism are well known to modify serum lipid and lipoprotein concentrations. The combined effect of OC use and apo E genotype on the concentration of apo E or apo C-III in apo B- (apo E-LpB or apo C-III-LpB) or in non-apo B-containing lipoparticles (apo E-Lp-non-B or apo C-III-Lp-non-B) are unknown. Our study comprised 613 women, aged 30-45 years, genotyped for common apo E polymorphism and who differed in their combined low-dose OC consumption. The concentrations of apo C-III, apo C-III-LpB and apo C-III-Lp-non-B were significantly higher in OC users than in non-users by 13, 23 and 8% respectively, without significant interaction with the apo E genotype. The concentrations of apo E and apo E-Lp-non-B were significantly lower (differences being -14% and -31% respectively) in OC users than in controls whereas the apo E-LpB concentration was significantly higher (+19%), resulting in a redistribution of apo E from Lp-non-B towards LpB. Total apo E and apo E-Lp-non-B concentrations were higher in subjects carrying the epsilon2 allele and lower in those with the epsilon4 allele when compared to epsilon3/epsilon3 subjects (P < 0.001). The opposite held for the apo E- LpB concentration (P < 0.05). The main finding is the significant interaction between apo E genotype and OC use (P < 0.01) on apo E-Lp-non-B concentration, the epsilon4 carriers showing the smallest differences between OC users and non-users in comparison with the epsilon2 or epsilon3/epsilon3 carriers. These results suggest that the common apo E polymorphism can modulate the OC use effect.

Lipoprotein lipase (C/G)447 polymorphism and blood pressure in the Stanislas Cohort

OBJECTIVE

Association between blood pressure and triglyceride levels, and between lipoprotein lipase (LPL) (C/G)447 polymorphism and triglyceride levels has been described. We investigated whether the LPL (C/G)447 polymorphism was associated with blood pressure (BP) levels and longitudinal changes.

DESIGN AND PARTICIPANTS

For cross-sectional analysis, 767 men and 816 women (29-55 years) were selected from the Stanislas Cohort, a cohort of volunteers for a free health check-up. Only subjects without anti-hypertensive or lipid-lowering medication were included in the study. A subset of this sample population, 359 men and 337 women, had been followed during the 11 years prior to recruitment in the Stanislas Cohort and was used for longitudinal analysis.

RESULTS

The cross-sectional study showed that serum triglyceride levels differed significantly according to LPL genotypes in both genders, the G447 allele being associated with the lowest triglyceride levels (P < or = 0.01). Univariate and multivariate analysis found that LPL polymorphism was not related to BP levels in men. In contrast, women with the LPL-G447 allele had lower systolic (SBP) and pulse (PP) pressure levels than those with the LPL-CC genotype (P < or = 0.01 and P < or = 0.05, respectively); this association being independent of triglyceride level. The longitudinal study showed LPL genotype was an independent predictor of PP and SBP follow-up levels in women; changes over 11 years being lower for LPL-G447 allele carriers (P < or = 0.05). These associations were independent of triglyceride level.

CONCLUSION

The LPL-G447 allele was found associated with lower PP and SBP independently of triglyceride level in women. This result suggests that the LPL gene may influence blood pressure.

Dietary intakes, eating style and overweight in the Stanislas Family Study

OBJECTIVES

To describe the eating patterns of members of French families and to assess the relationships between dietary intakes, eating style and overweight.

DESIGN

Cross-sectional analysis of nutritional and behavioural characteristics.

SUBJECTS

1,320 members of 387 families (age 11-65y) attending the Centre for Preventive Medicine for a routine medical check-up.

MEASUREMENTS

Individual body weight and height were measured. Food intake was assessed using a three day dietary record. Eating style was measured using the French validated version of the Dutch Eating Behaviour Questionnaire.

RESULTS

In each of the four groups (men, women, boys and girls), dietary restraint was positively correlated with overweight (P< or =0.001) and associated with lower energy intakes (P < or = 0.05-P < or = 0.001). A negative association between energy intake and overweight was found in girls only (P< or =0.001). In all cases, overweight and dietary restraint exaggerated any existing macronutrient imbalance in energy intake (ie higher protein and fat contributions, lower carbohydrate contribution). Emotional eating was positively correlated to body mass index in women only (P< or =0.01). External eating was mainly a characteristic of children (P< or =0.001).

CONCLUSION

As in overweight subjects, clear relationships were found in this sample of general population between dietary intakes and eating style. The population will be followed up for 10y. In the long term, these results should have implications in the prevention of obesity.

High sensitivity C-reactive protein: biological variations and reference limits

Serum C-reactive protein (CRP) concentration was determined for 3605 subjects using an immunonephelometric assay improved to provide greater sensitivity. Subjects were from 5 to 75 years old and belonging to 1003 nuclear families recruited from the Stanislas Cohort Study between January 1994 and August 1995. Sample values for CRP ranged from 0.17 mg/l to 100 mg/l. Geometric means (mean - SD; mean +/- SD) were in the 5-14 years old group 0.37 (0.17-1.07) mg/l, in the 15-28 years old group 0.47 (0.17-1.38) mg/l and in the 29-75 years old group 0.98 (0.34-2.85) mg/l. For women, the geometric means were 0.38 (0.17-1.10) mg/l, 0.62 (0.20-1.90) mg/l and 0.98 mg/l (0.31-3.13) mg/l respectively. The interindividual variability ranged from 138% to 759% among different age classes. Biological factors associated with CRP concentration variations were examined and accounted for 25% of the CRP variability in men and 40% in women. The main biological factors statistically associated with CRP concentration variations in men were: drugs, leukocyte count, body mass index, tobacco consumption, age, and in women: drugs, leukocyte count, age, body mass index and hemoglobin concentration. These factors were used to define the exclusion and partition criteria when obtaining the reference samples. Medians for reference values ranged from 0.20 to 0.68 mg/l in males and from 0.20 to 0.78 mg/l in women.

Association of the G-2548A polymorphism in the 5' region of the LEP gene with overweight

Mutations in the translated part of the leptin gene (LEP) have been found in only two families. Nevertheless DNA polymorphisms in the LEP region are linked to extreme obesity. We previously found in the 5' region of LEP a polymorphism, G-2548A, associated with a differerce in BMI reduction following a low calorie diet in overweight women. Recently, this polymorphism was associated with extreme obesity in women. In this work, we genotyped a new sample from the general population including 314 normal weight (BMI < 27 kg/m2) and 109 overweight subjects (BMI > or = 27 kg/m2). The genotype and allele frequencies were significantly different between groups, with the G-2548 allele being more frequent in the overweight subjects (p < 0.01). In men, carriers of this allele had lower leptin concentrations adjusted for fat mass (p = 0.05). Our results indicate that variations at the leptin locus are associated with common obesity phenotypes, and not only with extreme obesity or the rare mendelian obesity syndromes.

Apolipoprotein E polymorphisms and concentration in chronic diseases and drug responses

Apolipoprotein (apo) E is an important circulating and tissue protein involved in cholesterol homeostasis and many other functions. The common polymorphism in the coding region of the gene, four polymorphisms in the promoter region, other additional single nucleotide polymorphisms, as well as several apo E variants have been identified. The common coding polymorphism strongly influences the lipid metabolism and the circulating concentration of apo E itself. This polymorphism is at the origin of the implication of apo E in cardiovascular and neurodegenerative diseases, but also of the relation of apo E with longevity. Probably due to its many metabolic and functional consequences, apo E polymorphism has been shown to influence the responses of patients to several drugs (fibrates, statins, hormone replacement therapy, anti-Alzheimer drugs) or environmental interventions (black tea, alcohol, diet). Apo E genotyping may be clinically helpful in defining the risk of patients and their responses to therapeutics. Finally, circulating apo E concentration appears to be altered in diseases and can be modulated by some of the drugs cited above. This parameter can thus also give interesting clinical information and could be a therapeutic target, providing it is validated. At the present time, we cannot exclude that apo E concentration may be the most prominent apo E parameter to be considered in health and disease, while apo E polymorphisms would represent only secondary parameters influencing apo E concentration.

Apolipoprotein E polymorphism and serum concentration in Alzheimer's disease in nine European centres: the ApoEurope study. ApoEurope group

As part of the ApoEurope Project, apolipoprotein E (apo E) common polymorphism and serum concentration were determined in 489 Alzheimer's disease patients and 429 controls. Patients and controls were recruited through nine centres in eight European countries. Age, sex ratios and education levels of both case and control populations were similar, although discrete differences appeared between centres. The prevalence of the epsilon4 allele was higher in Alzheimer's disease than in controls (increased by 140%), while serum apo E concentration was lower by 11.2% (p<0.001). In addition, serum total cholesterol and triglyceride concentrations were lower in Alzheimer's disease (p<0.001), while that of apo Al was not affected. The decrease in serum apo E concentration was not accounted for by the epsilon4 allele, age or gender, suggesting that apo E concentration might represent an additional risk factor for Alzheimer's disease, complementary and independent of the epsilon4 allele. Further analysis will be aimed at determining whether the quantitative link between apo E concentration and Alzheimer's disease occurs through the effect of apo E genotype on lipid parameters or by other mechanisms.

5-HT2A receptor gene polymorphism is associated with food and alcohol intake in obese people

OBJECTIVE

To test the association between a polymorphism of the 5-HT2A receptor gene, -1438G/A, and energy and nutrients intake, including alcohol.

SUBJECTS

Two hundred and seventy six unrelated overweight subjects (180 women, 96 men) were recruited from the Nutrition Department of Bichat Hospital in Paris on the basis of 120% of ideal body weight (body mass index, BMI=33.3+/-4.8 kg/m2). A second overweight sample (31 women, 49 men) was drawn from the Stanislas Family Study, composed of volunteers for a free health examination in Nancy (BMI=29.6+/-3.1 kg/m2).

MEASUREMENTS

Energy and nutrients intake were assessed using the diet history method in Paris and the 3-day record method in Nancy. We analyzed the polymorphism by PCR followed by MspI digestion. Statistical differences between genotypes were assessed by using the non-parametric Kruskal-Wallis test.

RESULTS

In the whole overweight population, the A allele was associated with lower energy intake 10. 3+/-2.8, 9.9+/-2.8, 9.3+/-2.9 MJ/day for GG, GA and AA genotypes respectively (P<0.05). This association was significant in the patient sample from Paris and in the overweight male volunteers from Nancy. Allele A-related lowering in energy intake was due to a trend to lower intakes in all the main nutrients. The A allele was also associated with a lower alcohol consumption: 18.4+/-19.7, 15.3+/-21. 2 and 12.3+/-17.5 g/day for GG, GA and AA genotypes, respectively (P<0.05).

CONCLUSIONS

These data indicate that a gene polymorphism may influence food and alcohol intake in overweight humans. This could be explained by the role of the serotonergic system as a determinant of food intake.

Segregation analysis of fat mass and fat-free mass with age- and sex-dependent effects: the Stanislas Family Study

Segregation analysis using a regressive model with age- and sex-dependent effects was applied to family data of weight, fat mass (FM) and fat-free mass (FFM) to investigate the major gene hypothesis. The sample consisted of 220 nuclear families from the 'Stanislas Cohort' who volunteered for a free health examination (n = 913). FM and FFM were assessed by bioelectrical impedance. The data were adjusted for height2 and height prior to analysis. The spouse, father-offspring, mother-offspring and sib-sib correlations were: 0.16, 0.18, 0.25 and 0.32 for weight; 0.13, 0.20, 0.23 and 0.28 for FM; 0.18, 0.16, 0.29 and 0.41 for FFM. For the three phenotypes, models specifying a major gene with age- and sex-dependent effects and residual family correlations was better supported than models including only family correlations. For weight, the most parsimonious genetic model was a codominant model with a sex-dependent effect in parents and an age-increasing effect in offspring. For FM, the most parsimonious model was also a codominant model with sex-dependent effects in parents indicating higher effects in women than in men. For FFM, the most parsimonious model was a recessive model with no significant age or sex interaction, although the age interactions paralleled those observed on weight in offspring. For weight and FM, mendelian transmission was rejected. For FFM, the Mendelian and the environmental hypotheses were nearly equally supported and none was rejected when compared to general transmission. Then, evidence for a single major gene could not be inferred for any of the traits. This does not preclude the existence of several genes acting in a more complex way. However, our findings emphasize that weight is a composite phenotype reflecting different components which evolve in distinct ways during life span. For this reason, FM should be highly preferred to weight or BMI for the research of susceptibility genes to obesity.

Bivariate familial correlation analysis of quantitative traits by use of estimating equations: application to a familial analysis of the insulin resistance syndrome

Familial correlation analysis involving two traits may give a better insight into the etiology of multifactorial syndromes than familial analysis focused on single traits. Significant cross-trait correlations between biological relatives but not between spouses suggest that the two traits share common transmissible factors whereas correlations between spouses additionally suggest the influence of shared lifestyle factors. We apply the Estimating Equations (EE) technique to the estimation of intra-trait and cross-trait familial correlations on two quantitative traits. Unlike maximum likelihood methods, the EE method does not require one to specify the joint distribution of the traits. Estimation of correlations and of their variance involves an iterative three-stage algorithm which converges rapidly. The generalized Wald test can be used to test any specific hypothesis of familial resemblance. This method has great flexibility for handling covariates and incomplete family data. A simulation study indicated that the EE technique performed well in large samples (100 families), both in terms of type I error and coverage probability . However, in small samples (50 families), an increase of the type I error and a decrease of the coverage probability was observed. As an illustration, we applied this technique to a family study of metabolic factors involved in the Insulin Resistance Syndrome (body mass index, insulin, triglycerides, HDL-cholesterol, and diastolic blood pressure). The study was carried out in a sample of 216 healthy nuclear families with > or =2 offspring. The results suggested the existence of a common transmissible (genetic or cultural) factor influencing both body mass index and insulin, whereas the weak clustering of triglycerides and HDL-cholesterol would be more compatible with the influence of shared lifestyle factors.

Effect of short- and long-term storage on human serum and recombinant apolipoprotein E concentration

In order to assess the short- and long-term stability of apolipoprotein (apo) E concentration in serum, we compared the apo E concentrations measured in fresh human serum samples with those determined after storage at +4 degrees C, -20 degrees C or -80 degrees C. The serum apo E concentration was measured by immunoturbidimetry using an anti-human apo E polyclonal antibody from goats. One week storage at +4 degrees C did not significantly affect the serum apo E concentration. At -20 degrees C or -80 degrees C no significant change in apo E concentration occurred during up to three months of storage. Moreover, the concentration of apo E was not modified after long-term storage of serum samples kept at -196 degrees C in liquid nitrogen for up to four years. In addition, 15 freeze-thaw cycles, over a 3-week period, did not affect the apo E concentration in serum. A similar freeze-thaw procedure applied to purified human recombinant apo E showed that apo E2 isoform was the most stable in comparison with the apo E3 and apo E4 isoforms.

Measurement of thiols in human plasma using liquid chromatography with precolumn derivatization and fluorescence detection

A liquid chromatography (LC) method for the simultaneous measurement of the main low molecular mass thiols (i.e., cysteine, cysteinylglycine, homocysteine, and glutathione) in human plasma is described. The sample treatment consists of the reduction of disulfide bounds with tri-n-butylphosphine and protein precipitation with trichloroacetic acid followed by precolumn derivatization with a thiol-selective fluorogenic reagent (7-fluoro-2,1,3-benzoxadiazole-4-sulfonamide). The structure of thiol derivatives is assessed using electrospray ionization-mass spectrometry (MS). The stability of resulting adducts in acidic medium (24 h at 10 degrees C) allows the automation of the technique and a high throughput of samples (approximately 50 per day). Separation is complete within 12 min using isocratic reversed-phase mode, and detection is operated by spectrofluorimetry (lambda ex = 385 nm and lambda em = 515 nm). Quantitation is performed by an internal standardization mode using thioglycolic acid. The LC method is fully validated, and homocysteine concentrations obtained in plasma samples are compared with values measured using either fluorescence polarization immunoassay or capillary gas chromatography-MS; a good correlation is observed between LC and both methods. The method has been applied in daily use to a large-scale study in a human healthy population, and some resulting data are discussed.

Characterization and quantification of serum lipoprotein subfractions by capillary isotachophoresis: relationships with lipid, apolipoprotein, and lipoprotein levels

Human serum lipoproteins are currently defined according to their density as well as according to their electrophoretic mobility. They can be fractionated into discrete subspecies which exhibit variations in their structure and function. Capillary electrophoresis has been suggested to be a potential analytical strategy in understanding metabolic lipoprotein heterogeneity. In a sample of 35 normolipidemic subjects, we analyzed ceramide-labeled serum lipoproteins by capillary isotachophoresis linked to laser-induced fluorescent detection. Capillary isotachophoresis showed advantage to be an automated, rapid (6 min) and reproducible (CV < 7%) separation mode, on-line monitoring lipoprotein subfractions according to net charge. HDL were separated into three subfractions: i) the fast migrating HDL correlated positively with serum apoA-I (P < 0.05) and negatively with triglyceride (P < 0.01) concentrations, ii) the intermediate migrating HDL involved in HDL-cholesterol delivery and inversely related to LDL particles concentration (P < 0.001), and iii) the slow migrating prebeta(1)HDL. Triglyceride level was significantly associated with two fractions: i) the VLDL fraction correlated positively with apoE serum concentration (P < 0.01), and ii) the IDL fraction closely and positively associated with apoC-III-containing lipoprotein level (P < 0.001). Two LDL subfractions were positively related to LDL-cholesterol (0.05 </= P < 0.01) and might characterize, respectively, small dense and large buoyant LDL subfractions: i) the fast migrating LDL, positively linked to apoB concentration and to LpCIII:B (P < 0.01) reflecting altered IDL metabolism, and ii) the slow migrating LDL. Analytical capillary isotachophoresis of fluorescent-stained lipoprotein subfractions might represent an efficient qualitative and quantitative tool which would afford complementary information on lipoprotein metabolism to current clinical lipoprotein analysis.

Intima-media thickness and diameter of carotid and femoral arteries in children, adolescents and adults from the Stanislas cohort: effect of age, sex, anthropometry and blood pressure

OBJECTIVES

To study carotid and femoral intima-media thicknesses and diameters in relation to age, sex, morphologic status and blood pressure.

PARTICIPANTS

The subjects were 369 men and women (aged 10-54 years) from the Stanislas cohort, with no known cardiovascular disease.

METHODS

Intima-media thicknesses and diameters were measured by B-mode ultrasonography. The effects of sex, age, smoking, anthropometric variables, cholesterol and blood pressure were studied using bivariate and regression analysis.

RESULTS

Carotid and femoral intima-media thicknesses were not affected by age nor by sex up to 18 years of age. Thereafter, they increased sharply in men and remained higher than in women. Values were correlated with systolic blood pressure only in men, and with fat-free mass in children and young adults only at the femoral site. Smoking, body mass index and fat mass were associated with intima-media thicknesses only in adults. Carotid diameter was little affected by age during childhood and in adults. Femoral diameter increased up to the age of 18 in both sexes and remained unaffected by age thereafter. This increase was more pronounced in boys, and so values became consistently greater in males aged over 14 years. Carotid diameter was correlated with body mass index or fat mass whereas femoral diameter was correlated with weight or fat-free-mass in children and men. The opposite was observed in women.

CONCLUSION

Sex differences occur before adolescence for arterial diameter, but only at an adult age for intima-media thickness. In young subjects, carotid geometry seems to be influenced by blood pressure and excess body weight, while femoral artery geometry seems to be related to blood pressure and body growth.

Apolipoprotein E4, lipoprotein lipase C447 and angiotensin-I converting enzyme deletion alleles were not associated with increased wall thickness of carotid and femoral arteries in healthy subjects from the Stanislas cohort

Studies have shown contrasting results concerning the relation between carotid intima-media thickness (IMT) and apolipoprotein E (apo E) and angiotensin-converting enzyme (ACE) polymorphisms. Subjects, 76 men and 74 women, between 33 and 50 years, without any history of cardiovascular disease and without any anti-hypertensive or lipid lowering medication were selected from the Stanislas cohort. The IMT of carotid and femoral arteries were investigated by B-mode ultrasonography. The common apo E, (C/G)447 lipoprotein lipase (LPL) and I/D ACE gene polymorphisms and serum ACE activity were determined. In the overall sample, male sex, age, systolic blood pressure, BMI, serum apo B level and tobacco consumption were positively correlated with carotid and femoral IMT. The common apo E polymorphism, the (C/G)LPL447 polymorphism and ACE activity were not related to carotid and femoral IMT variability in either men or women. Unexpectedly, the I allele of the ACE gene was related to higher femoral IMT than the D allele in non-smokers only. Similar results were observed after adjustment for the main covariates of IMT variability. In conclusion, amongst our young adult sample the candidate risk factors for cardiovascular disease, apo epsilon4, C447-LPL and D-ACE alleles and ACE activity were not associated with increased carotid and femoral IMT.

Genetics strongly determines the wall thickness of the left and right carotid arteries

In 76 supposedly healthy families, we investigated the familial resemblance of left and right carotid intima-media thickness (IMT) measured by B-mode ultrasonography and the impact of the common apolipoprotein E (apo E) polymorphism and the insertion/deletion polymorphism of the angiotensin-converting enzyme (ACE). Genetic factors accounted for about 30% of IMT variation. The insertion/deletion ACE polymorphism did not influence carotid IMT, whereas apoE polymorphism explained about 1.5% of only right carotid IMT variability independently of cholesterol levels. The apo epsilon2 and apo epsilon4 alleles were associated with lower right carotid IMT than was the apo epsilon3 allele. We conclude that genetic factors strongly contribute to IMT variability in healthy people and that the apo E polymorphism may be one of these factors.

Effects of supplementation with a combination of antioxidant vitamins and trace elements, at nutritional doses, on biochemical indicators and markers of the antioxidant system in adult subjects

OBJECTIVE

To test the impact of supplementation with nutritional doses of antioxidant nutrients on biochemical indicators of vitamin and trace element levels.

DESIGN

A randomized double-blind trial was performed comparing two groups receiving daily either a combination of vitamins (beta-carotene, 6 mg; vitamin C, 120 mg; and vitamin E, 30 mg) and trace elements (zinc, 20 mg; and selenium, 100 micrograms); or a placebo.

SUBJECTS

401 subjects (166 males aged 45 to 60 years and 235 females aged to 35 to 60 years). MEASURE OF OUTCOME: Biological markers of vitamin and trace element status and free radical parameters were measured initially, 3 months, and 6 months after supplemention.

RESULTS

Mean serum concentrations of alpha-tocopherol, vitamin C, beta-carotene, zinc and selenium increased significantly after 3 months of supplementation in the group receiving multivitamins associated with minerals. At baseline, 18.2% of the men and 5.1% of the women had low concentrations of serum vitamin C (< 20 mumol/l): 2.4% of the men and 17% of the women presented low concentrations of serum retinol (< 1.4 mumol/l): 18.7% of men and 10% of women had serum beta-carotene < 0.30 mumol/l. None of the study subjects had serum alpha-tocopherol concentrations below the limit cut-off point (< 9.3 mumol/l). Low serum zinc concentrations (< 10.7 mumol/l) were found in 15.1% of men and 23.8% of women. Low serum selenium concentrations (< 0.75 mumol/l) were found in 6% of men and 6.4% of women. A significant increase in plasma and red cell GPx activity was observed in groups receiving supplementation. No modifications were observed after 6 months of supplementation for malondyaldehyde.

CONCLUSION

This study demonstrates the efficacy of an intake of antioxidant vitamins and trace elements, given at nutritional doses, on biochemical indicators of vitamin and trace elements status.

[Criteria for addition of apolipoprotein B measurement during routine health screening examinations: validation in a Stanislas cohort. The Biologist's Group of the Health Examination Center]

The purpose of this study was to determine the criteria in which apolipoproteins AI and B should be performed during periodic health screening examinations. Clinically, the results of apolipoproteins AI and B are most useful when there are minor lipid perturbations (cholesterol and triglycerides), but their routine determination is not justified. Several mathematical models, defined by discriminate factorial analysis, have been studied. The one based uniquely on cholesterol and triglyceride concentrations was the most efficient in identifying patients in whom apolipoprotein B should be determined. In contrast, no model was found for the indication of apolipoprotein AI determination. Data from a population different from that used for the model establishment were used to validate the model. This strategy, determining criteria for additional measurement of apolipoprotein B in a general population, is of clinical interest because it may provide complementary information for subjects with atherogenic risk. Moreover, it is of economical interest because it has the potential of limiting complementary testing and its associated costs.

Elevated lipoprotein(a) levels and small apo(a) isoforms are compatible with longevity: evidence from a large population of French centenarians

Epidemiological studies have shown lipoprotein(a) (Lp(a)) to be an independent risk factor for cardiovascular disease. Lp(a) is a cholesterol-rich, low-density lipoprotein (LDL)-like particle to which a large glycoprotein, apolipoprotein(a) (apo(a)) is attached. Plasma Lp(a) levels are highly genetically determined and influenced to a minor degree by environmental factors. In an effort to determine whether Lp(a) might be associated with longevity, we have evaluated Lp(a) levels and apo(a) isoform sizes in a population of French centenarians (n = 109) compared to a control group (n = 227). The mean age of centenarians was 101.5 +/- 2.4 years while the control group was 39.4 +/- 7.2 years. Plasma levels of total cholesterol and triglyceride were within the normal range in both centenarian and control subjects. Lp(a) levels were higher in centenarians (both male and female) than in the normolipidemic control group (mean Lp(a) level = 0.33 +/- 0.42 and 0.22 +/- 0.27 mg/ml, respectively, P < 0.005). The distribution of apo(a) isoforms was significantly shifted towards small isoform size in the centenarian population as compared to the controls (54.4 and 41.4% of isoforms < or = 27 kringles (kr), respectively, P = 0.04). Nonetheless, the apo(a) size distribution in centenarians did not entirely explain the high Lp(a) levels observed in this population. Factors other than apo(a) size, and which may be either genetic or environmental in nature, appear to contribute to the elevated plasma Lp(a) levels of our centenarian population. We conclude therefore that high plasma Lp(a) levels are compatible with longevity.

Diet, antioxidant status, and smoking habits in French men

The aim of this study was to assess the association between smoking, food consumption, and antioxidant vitamin intake and plasma indexes of oxidative stress and antioxidant defenses in French adults. Food and nutrient intakes of 459 healthy men aged 23-57 y were estimated by the diet history method and analyzed by smoking status. Plasma alpha-tocopherol, ascorbic acid, and carotenoids were measured as antioxidants and malondialdehyde, protein Schiff bases, and autoantibodies against malondialdehyde-protein adducts as oxidative stress indexes. Smokers ate less fruit and vegetables than nonsmokers, leading to lower vitamin E, vitamin C, and carotene intakes, even after adjustment for age, education, and marital status. Unlike vitamin E, plasma ascorbic acid and beta-carotene concentrations were reduced in smokers compared with nonsmokers and were inversely related to cigarette consumption. This difference remained significant after adjustment for alcohol and dietary intakes. Among the measured oxidative stress indexes, only Schiff base concentration was positively related to the number of cigarettes smoked. In our sample of French men, smoking had an adverse effect on antioxidant status; vitamin intakes were reduced in smokers and plasma antioxidant indexes were altered independently of dietary intakes. As in other countries, in France smokers require particular attention in terms of public health intervention.

Objectives, design and recruitment of a familial and longitudinal cohort for studying gene-environment interactions in the field of cardiovascular risk: the Stanislas cohort

The main objective of the Stanislas cohort is to study the role and the contribution of genetic and environmental factors to cardiovascular status. We plan: a) to describe the degree of association of a large number of cardiovascular risk indicators with cardiovascular endpoints, b) to evaluate the contribution of genetic and that of environmental factors to this association, c) to follow the evolution of these risk indicators during a period of at least ten years, d) to search for the determinants influencing this evolution. The principal variables studied are: a) blood pressure, cardiac mass, and wall thickness of carotid and femoral arteries, b) obesity and fat mass, c) indicators of lipid metabolism, d) genetic polymorphisms of several cardiovascular risk candidate genes, e) food, tobacco and alcohol consumption, f) consumption of drugs and anti-oxidant vitamins. Between September 1993 and August 1995, 1006 families consisting of the two biological parents with at least two children were recruited totalling 4295 individuals. This cohort will be followed up until 2004. There will be two health examinations five and ten years after the initial examination. A bank of blood samples (serum and plasma) in liquid nitrogen and DNA (-80 degrees C) has been established.

Low and very low density lipoprotein composition and resistance to copper-induced oxidation are not notably modified in smokers

To study whether tobacco use was associated with oxidative phenomena affecting lipoproteins, we estimated susceptibility of LDL and VLDL to an in vitro copper-mediated oxidation, and measured serum autoantibody titers against oxidized LDL in 45 middle-age healthy nonsmokers, 35 smokers and 37 ex-smokers of both sexes, taking into account the detailed lipid composition of the lipoproteins. VLDL from female smokers had higher triglyceride, phospholipid, apolipoprotein E and alpha-tocopherol content and showed a higher rate of copper-induced oxidation in comparison with those from nonsmokers (P < or = 0.05) whereas the relative composition of these particles in saturated, mono- or poly-unsaturated fatty acids was not modified by tobacco consumption. After adjustment for triglyceride content, no statistically significant difference in oxidation rate was observed. Lipid, alpha-tocopherol and protein composition of LDL did not appear to be influenced by smoking; in accordance with these observations, no difference in indices of in vitro oxidizability of LDL was noticed between the different groups. Autoantibody titers against oxLDL were similar in smokers and nonsmokers. We conclude that, in supposed healthy individuals, smoking does not seem to be associated with notable variations in composition of VLDL and LDL or with an increase of oxidizability of these atherogenic lipoproteins.

Measurement methods of drug consumption as a secondary judgment criterion for clinical trials in chronic rheumatic diseases

Drug consumption is sometimes used as a secondary judgment criterion for clinical trials. Many measurement methods are available to quantify drug consumption. Several methods were applied in a rheumatic disease trial involving 121 patients with chronic low back pain who lived around Saint-Nectaire, France, and who participated in the trial from April to November 1993 to determine an easily used and practical measurement method to detect a significant drug consumption change over time. Analgesic and nonsteroidal antiinflammatory drugs (NSAIDs) were classified according to the anatomical therapeutic chemical classification. Consumption was quantified on a weekly basis in number of tablets (method 1), unit of defined daily dose (method 2), milligrams of active principle (method 3), and NSAID equivalence score (method 4). These methods were applied in a randomized clinical trial of spa therapy on sufferers of chronic low back pain. An analysis of variance with repeated measures showed a significant difference in drug consumption between treatment and control groups detected by all methods, except for the NSAID consumption measured with method 3. The comparison of each method by the relative efficiency index indicated that method 1 had a greater sensitivity for detecting changes of drug consumption. Tablet count appears to be a more sensitive and more practical method for detecting a drug consumption change in clinical trials.

Cerebrospinal fluid apolipoprotein E level is increased in late-onset Alzheimer's disease

Worldwide evidence has recently shown that the allele epsilon4 of apolipoprotein E (ApoE) is a genetic risk factor for Alzheimer's disease (AD) underlining the possible role of apoE in the physiopathology of AD. To evaluate the usefulness of apoE concentration in pathogenesis of AD, we measured the cerebrospinal fluid (CSF) levels of apoE. CSF apoE level was significantly higher in 38 patients with late-onset AD than in 31 control patients and 47 patients suffering from other neurological and related diseases. Higher levels of CSF apoE were also present in a subset of patients with meningoencephalitis, motor neuron disease, and low back pain. The increase of CSF-apoE in AD is in agreement with results from studies that find an increase of mRNA apoE in the brains of AD patients. Compared to other works, these results underline the importance and the difficulties of the selection of the controls. The CSF apoE level seems to be a reflection of neuronal damage and/or an inflammatory reaction that may be common to AD and other neurological and related diseases.

Changes in serum apolipoprotein and lipoprotein profile induced by chronic alcohol consumption and withdrawal: determinant effect on heart disease?

The effects of alcohol consumption on serum concentrations of apolipoproteins (apo) A-I, C-III, B, and E and of lipoproteins (Lp) A-I, A-I:A-II, C-III, C-III:B, and (a) were studied in 132 healthy subjects, including 55 low drinkers of alcohol (&lt;20 g/day), 36 moderate drinkers (20-50 g/day), and 41 heavy drinkers (&gt;50 g/day), and in 97 hospitalized alcoholic patients (&gt; 100 g/day) without severe liver disease (especially functional insufficiency), before and after 21 days of withdrawal treatment. Serum concentrations of apo A-I, LpA-I, LpA-I:A-II, apo C-III, and LpC-III significantly (P &lt;/= 0.01) increased with alcohol intake (mean +/- SE in low drinkers vs in alcoholics)--1.45 +/- 0.03 vs 1.78 +/- 0.05 g/L; 0.45 +/- 0.02 vs 0.56 +/- 0.02 g/L; 0.99 +/- 0.02 vs 1.22 +/- 0.04 g/L; 27.6 +/- 1.5 vs 39.7 +/- 1.7 mg/L; and 8.4 +/- 0.9 vs 24.7 +/- 1.7 mg/L, respectively-whereas apo B and LpC-III:B concentrations tended to decrease--1.20 +/- 0.04 vs 1.06 +/- 0.04 g/L and 19.3 +/- 1.2 vs 14.9 +/- 1.0 mg/L, respectively. No significant difference between these four types of alcohol consumption was noticed for cholesterol, triglycerides, apo E, and Lp(a). After withdrawal, the concentrations of serum apo A-I, apo C-III, LpA-I, LpA-I:A-II, and LpC-III decreased significantly (P &lt;/= 0.01), reaching values comparable with those in low drinkers; concentrations of triglycerides, apo B, apo E, and Lp(a) rose; and cholesterol concentration was unaffected. In multiple regression analysis, after adjustment for serum concentrations of albumin, aspartate aminotransferase, and gamma-glutamyltransferase and for the Quetelet index, alcohol consumption remained positively correlated to apo A-I, LpA-I:A-II, apo C-III, and LpC-III concentrations. Study of other determinants of serum apo and lipoprotein concentrations suggests that alcohol-related variations in some of them, especially apo A-I, might depend on the metabolic ability of the liver to synthesize proteins and on induction phenomena. Finally, although the increase of antiatherogenic apo- and lipoproteins and the decrease of those known to be atherogenic were generally marked in alcoholics, alcohol-related modifications of these markers were very limited in our sample of French healthy men. We conclude, therefore, that moderate alcohol consumption (20-50 g/day) is unlikely to protect against ischemic heart disease through an effect on the proteins measured in this study.

Family resemblance in energy and macronutrient intakes: the Stanislas Family Study

BACKGROUND

There seems to be a consensus that family influences on dietary habits are important but few studies have addressed this issue directly. The purpose of this study was to determine if and how dietary intake aggregates within families.

METHODS

We examined the family aggregation of energy intake and the proportion of protein, fat and carbohydrate in the diet, estimated by a 3-day food consumption diary in 387 middle-class French families.

RESULTS

For energy and all macronutrients, spouse-spouse and child-child correlations were higher than parent-child correlations suggesting the minor contribution of genetics and the preponderant role of cultural and residual random environment. Variance component analysis confirmed the absence of genetic component for energy and all macronutrients and underlined the important role of a cohabitational effect for parents. Cultural inheritance represented 30-40% of dietary intake variance for children. Families who shared meals together more often had a lower residual random component. With the increasing number of meals eaten together (> 45/week versus < or = 45/week), between-generation components increased by about 10% for fat and carbohydrate, while for protein intake, the between-generation component for both parents (about 27%) and children (about 37%) remained unchanged.

CONCLUSIONS

The general finding that dietary intake aggregates within families and that the individual behaviours are greatly influenced by characteristics within the family unit such as the number of meals eaten together provides additional justification for health promotion programmes that target the family as the unit for intervention.