点击显示 收起
1 From INSERM U525, Epidemiologic and Molecular Genetics of Cardiovascular Diseases, Faculté de Pharmacie, Nancy, France (BH, EA, FC, GS, and SV); INSERM U695, Genetics of Type 2 Diabetes and of its Vascular Complications, Faculté Xavier Bichat, Paris, France (FF and RA), and the Centre de Médecine Préventive, Vandoeuvre-lès-Nancy, France (GS)
2 Supported by a grant from the Institut Danone (Paris, France). 3 Reprints not available. Address correspondence to B Herbeth, INSERM U525, Faculté de Pharmacie, 30 rue Lionnois, 54000 Nancy, France. E-mail: bernard.herbeth{at}nancy.inserm.fr.
ABSTRACT
Background: Serotonin (5-hydroxytryptamine; 5-HT) is a key mediator in the control of food intake and is probably involved in the etiology of anorexia nervosa. An association between a polymorphism of the 5-HT receptor (5-HT2A) gene promoter (–1438G/A) and anorexia nervosa has been reported.
Objective: We investigated the relation between the –1438G/A polymorphism of the 5-HT2A gene and the energy and macronutrient intakes of children and adolescents.
Design: This cross-sectional study included 370 children and adolescents aged 10–20 y (176 boys and 194 girls from 251 families) drawn from the Stanislas Family Study. Energy and macronutrient intakes were assessed by using 3-d food records. The –1438G/A polymorphism was analyzed by polymerase chain reaction and then by Hpa II digestion.
Results: In the overall group, after adjustment for age, sex, weight, height, and family correlation, the A allele was significantly associated with lower energy (P for trend = 0.045) and with total, monounsaturated, and saturated fat intakes expressed in g/d (P for trend = 0.007, 0.005, and 0.006, respectively). Subjects with the GA genotype had intermediate values. In addition, genotype x sex and genotype x age interactions were not significant.
Conclusions: The 5-HT2A gene polymorphism in the promoter region is associated with energy and fat intakes in young people. This could be explained by the role of the serotonergic system as a determinant of food intakes and eating behavior.
Key Words: Serotonin • 5-hydroxytryptamine • 5-HT receptor • 5-HT2A • genetic polymorphism • food intake • children • adolescents
INTRODUCTION
Serotonin (5-hydroxytryptamine; 5-HT) is a key mediator in the control of food intake. Serotonin reduces food intake and is probably involved in the etiology of anorexia nervosa and in weight regulation (1); moreover, drugs that increase serotonergic transmission exert anorectic activity. An association between a polymorphism of the 5-HT receptor (5-HT2A) gene promoter (–1438G/A) and anorexia nervosa has been reported (2). In a previous study, we showed that this polymorphism could influence food and alcohol intakes in an adult overweight population because the A allele is associated with significantly lower energy and alcohol intakes and with a trend to lower intakes of proteins, fats, and carbohydrates (3). Because anorexia nervosa is almost exclusively found in young people and because we wanted to ascertain the validity of our previous results, we investigated in the current study whether the 5-HT2A polymorphism could influence energy and nutrient intakes in a sample of children and adolescents.
SUBJECTS AND METHODS
Subjects
This work is part of the Stanislas Family Study, a 10-y longitudinal follow-up study begun in 1994 in 1006 families selected at the Center for Preventive Medicine (Vandoeuvre-lès-Nancy, France) on the basis of being free of chronic or acute disease that could influence nutritional status (4). In this report, we present data obtained from a random subsample of 370 children and adolescents aged 10–20 y (176 boys and 194 girls from 251 families) at the first examination (1994–1995). All subjects underwent a complete medical examination including weight and height measurements. Body weight index was calculated as the ratio of weight to weight reference values for sex and age as delineated by Rolland-Cachera et al (5).
Written informed consent was obtained from each participant. The research protocol was approved by the Ethics Committee of Lorraine.
Methods
Dietary intake was assessed with the use of a 3-d dietary record (6), which was completed during 2 weekdays and 1 weekend day assigned at random for each family. All subjects received guidance from a dietitian on the procedures for completing the dietary record and measuring food portions. For young children, the 3-d diary was filled in by the mother and the child together. One week later, the 3-d record was checked and completed by the dietitian by using colored photographs of foods, each with 3 different portion sizes. Macronutrient and micronutrient intakes were estimated with a updated computerized version of the General Index of Foods (7).
Blood samples were collected after an overnight fast. Genomic DNA was extracted from peripheral blood samples by using the salting-out method (8). The presence of the –1438G/A polymorphism in the promoter region of the 5-HT2A gene was ascertained by polymerase chain reaction amplification with 2 primers: 5'-AAGCTGCAAGGTAGCAACAGC-3' and 5'-AACCAACTTATTTCCTACCAC-3'. The 468-base pair (bp) product was digested with Hpa II (which cuts the –1438 allele into 2 fragments of 244 and 224 bp), and the alleles were separated in an 8% polyacrylamide gel (9–11).
Statistical analysis
Statistical analyses were performed by using SAS software (version 8.01; SAS Institute Inc, Cary, NC). A chi-square test was used to determine whether the genotype frequency was in Hardy-Weinberg equilibrium. Multiple regression analysis was used to compare food intakes as a function of the 5-HT2A genotypes. To control for variables that could be related to energy and macronutrient intake, we used sex, age, height, and weight as covariates. In addition, sex x genotype and age x genotype interactions were evaluated. Because no interaction was significant and because we wanted to improve the statistical power, data from males and females were analyzed together. To calculate P for trend, the 3 genotypes were coded as a single variable and were assigned the following values: GG = 0, GA = 0.5, and AA = 1. Because persons within a family were not independent, regression analysis were conducted by using the SAS GENMOD procedure with a family factor as repeated statement. GENMOD was based on the generalized estimating equation, which provides a practical method with reasonable statistical efficiency to analyze correlated data such as familial data by modeling the covariance structure of the correlated measurements. Unlike other methods, the measurements must not be assumed to be multivariate normal (12). P 0.05 was accepted as significant.
RESULTS
In both males and females, the distribution of the –1438G/A polymorphism of the 5-HT2A gene did not significantly deviate from Hardy-Weinberg equilibrium. Food intakes and characteristics according to the 5-HT2A gene polymorphism in the entire group of boys and girls are shown in Table 1. Sex ratio, age, weight, height, body mass index, and body weight index did not differ significantly among the 3 genotypes. Because no significant sex x genotype interaction was found, only results in the overall sample of males and females were described. After adjustment for age, sex, weight, and height, the A allele was significantly associated with lower energy (P for trend = 0.045) and with total, monounsaturated, and saturated fat intakes expressed in g/d (P for trend = 0.007, 0.005, and 0.006, respectively); the values for subjects with the GA genotype were intermediate. Relative contributions of saturated, monounsaturated, and polyunsaturated fatty acids to fat intakes did not differ significantly between the 3 genotypes. When expressed as a percentage of daily energy levels, fat intakes were lower in A allele carriers than in G allele carriers (P for trend = 0.027).
View this table:
TABLE 1. Food intake and other characteristics according to the –1438G/A polymorphism of the serotonin receptor (5-HT2A) gene in children and adolescents aged 10–20 y1
DISCUSSION
In this study, we showed that the –1438G/A 5-HT2A polymorphism was associated with energy and fat intakes in a young population; A allele carriers had significantly lower intakes than did G allele carriers without significant relation to maturation indexes such as weight, height, and body mass index.
The genetic component of total energy and macronutrient intakes has been already documented by familial correlation and twin studies (13–15). Nevertheless, such investigations yielded a large range of heritability indexes, as reviewed by Faith and Keller (13). Moreover, in these studies, it can be difficult to separate true genetic effects from shared environment effects. For example, by using an extended family design including multiple family members with varied degrees of relatedness (ie, spouses, parent-child, siblings, and monozygotic and dizygotic twins), Pérusse et al (16) found significant genetic influences on macronutrient intakes (11% for protein and 20% for fat and carbohydrate) but not on total energy intake. One study showed that the familial effect could be adjusted for by more frequent intertwin contact among monozygotic twin pairs than among dizygotic twin pairs (17). Nevertheless, the results of a study of twins reared apart (18) yielded still higher correlations among monozygotic than dizygotic twins, which suggested that 20–30% of the variance in total energy and macronutrient intake could be determined by genes. In the San Antonio Family Heart Study (19), the familial relations accounted for 13–26% of the total population variation in total energy, fat, and carbohydrate intakes. In that study, familial influences were stronger when modeled as a genetic heritability than as a shared household effect. In a twin study in which nutrient intakes were measured directly instead of self-reported (20), the genetic component of total energy intake was estimated to be 24–33%, but the familial influences on specific macronutrient intakes could not be separated into genetic and shared environment components. In contrast, after quantifying eating behavior by using 7-d food diaries, de Castro (21, 22) observed significant genetic effects on daily intakes of total energy (65%), carbohydrate (8%), fat (10%), and protein (7%).
Including 2 publications from our group, only a few reports (3, 23, 24) have shown that a gene polymorphism may partly determine food intakes in humans. Two recent genome scans found chromosomal regions with some evidence of linkage for energy and nutrient intake, but not in the vicinity of the 5-HT2A receptor locus (25, 26). In our previous study in adults, the A allele was significantly associated with lower energy intakes, and associations with total, monounsaturated, and saturated fats were of borderline significance (3).
In the literature, the –1438A allele is significantly associated, according to a meta-analysis of case-control studies, with anorexia nervosa (2). This association is consistent with our data, which showed that the relation of genetic variation at the 5-HT2A receptor locus with food intakes can be evidenced also in a general population. A specific role of 5-HT2A in the regulation of food intake was shown in animal models, where specific agonists of 5-HT2A decreased the neuropeptide Y–stimulated food intake (27) and the 5-HT–induced hypophagia was antagonized by 5-HT2A receptor antagonists (28).
The 5-HT2A genotype was found to be significantly related to fat consumption. In human studies, there is evidence that some 5-HT drugs can readily reduce the intake of high-fat foods and data also suggest that 5-HT activation could lead to a selective avoidance of fat in the diet (29). Anorexia nervosa is prevalent mainly in young girls, and fat avoidance is a characteristic of the disease. Both these characteristics are in agreement with our results in young people of lower fat intakes associated with the –1438A allele. The associations of 5-HT2A genotype with food intakes were similar in boys and girls in our study, which could seem paradoxical, but, in a familial study of males with anorexia nervosa, it was shown that familial and genetic factors do not allow for distinction of the occurrence of anorexia nervosa between the sexes (30).
Conversely, the –1438G allele has been associated with abdominal obesity and with perturbations in the cortisol response to dexamethasone (31), and in the large National Heart, Lung, and Blood Institute's Family Heart Study genome scan, evidence of linkage was found between body mass index and the D13S257 genetic marker located in the same chromosomal region (13q14) as the 5-HT2A gene (32). Because fat intake has also been related to obesity, these results are consistent with those of the current study.
Nevertheless, we cannot exclude the possibility that the association of the –1438A allele with low energy and fat intakes may be due to linkage disequilibrium with other genetic polymorphisms that could affect eating behavior. However, apart from the results of our previous study of the same polymorphism in adults (3), no candidate gene other than the 5-HT2A gene has been shown to be related to food intakes in human. Another restriction is that the –1438G/A polymorphism could be also associated with personality or behavior traits that were related to food intakes or to reliance on self-reported food ingestion. Despite the scarcity of data in the literature, it may be hypothesized that persons carrying the –1438G/A polymorphism have altered mood or other personality characteristics that could be associated with food intake (33–35).
In most preliminary studies, such as this one, very little information about the genetic effect size is available beforehand, and thus it is difficult to calculate a reasonable sample size. According to power analysis based on the results of Aubert et al (3) in obese adults, in population with allele frequencies of 0.50, the number of the participants in the current study was adequate to evaluate 2.3–6% of heritability (ie, the percentage of variance that is explained by the polymorphism) with an additive mode of inheritance, achieving a statistical power of >80% at 5% probability (P value). In the sample of 371 children and adolescents, values for fat intake heritability were between 2.3% and 2.6%. At this time, we cannot rule out the possibility that the absence of association with intakes of the other macronutrients results from the low power due to the small sample size that was estimated on the basis of our former study with higher heritability. In addition, reliance on the 3-d dietary record could increase the random variability of diet intake and, consequently, the global variance of the studied traits. In overweight young subjects, for example, the food intake assessment could be less reliable (36, 37).
In conclusion, in spite of the limitations listed above, these data show that a gene polymorphism of the serotonergic system could be related to food intakes and eating behavior in humans. However, independent, confirmatory studies that involved larger samples with characteristics different from those in the children in the current study are necessary for definite conclusions about the role of this polymorphism in the physiology of eating behavior.
ACKNOWLEDGMENTS
We are deeply grateful for the cooperation of the families participating in the Stanislas Family Study cohort. We acknowledge the management, reception, preclinical, laboratory, and medical staffs of the Center for Preventive Medicine, Vandoeuvre-lès-Nancy, France. We especially thank Sylvie Péchiné for collection of food intake data, Maryvonne Chaussard and Chantal Lafaurie for family recruitment, and Dominique Aguillon for technical assistance with the polymorphism determination. Frozen aliquots of DNA were stored in the tissue bank at the Center for Preventive Medicine and at the INSERM U525 (Nancy, France).
BH, FF, and RA designed the study, performed the statistical analysis, and wrote the manuscript. EA and FC were responsible for laboratory analyses and participated in writing the manuscript. GS and SV are the principal investigators of the Stanislas Family Study and participated in writing the manuscript. None of the authors had any financial or personal conflict of interest.
REFERENCES