Objective: Eating disorders are conditions defined by abnormal eating habits that may involve either insufficient or excessive food intake to the detriment of an individual’s physical and mental health, which might be associated with the gynecological endocrinological conditions (amenorrhea, infertility) in young women. Some studies have revealed that there were a relationship between 5-HTTLPR gene and eating disorders. However, no clear conclusion has been offered addressing this issue. Therefore, we investigated the relationship of the 5-HTTLPR polymorphisms and eating disorders by meta-analysis.

Methods: PubMed, Medline, Psych INFO of web knowledge, and CNKI (Chinese National Knowledge Infrastructure) were searched before April 2012. Eighteen studies involving 2567 subjects. Data were entered into the Cochrane Collaboration review manager software (RevMan version 5). Individual and pooled 95% confidence intervals (CIs) were calculated.

Results: The results showed that the 5-HTTLPR L polymorphism was associated with significantly less risk of eating disorders (EDs) (including EDs, anorexia nervosa (AN) and Bulimia nervosa (BN)) [LL vs. SS (homozygote model), EDs: odds ratio (OR) =0.53, confidence interval (CI): 0.36-0.78; LL vs. SS, BN: OR=0.25, CI=0.11-0.57; LS vs. SS (heterozygote), AN, Asians: OR= 0.62, 95% confidence interval (CI): 0.46-0.85; LS vs. SS (heterozygote), BN: OR= 0.56, 95% CI: 0.39-0.79].

Conclusions: These results suggested that the 5-HTTLPR gene polymorphisms might be related with EDs under some conditions.

Chen W, Pu D, Ge H, Wu J (2014) 5-HTTLPR L Allele Gene may be a Protecting Factor for Eating Disorder -- A Meta-Analysis. Med J Obstet Gynecol 2(3): 1040.

• 5-HTTLPR
• Polymorphism
• Eating disorder
• Anorexia nervosa
• Bulimia nervosa
• Bing eating disorder 

Eating disorders (ED) form a series of Behavioral disorders diagnosed in young women (the age of onset is 16-17 years (rarely >30 years)) who share some syndromes associated with physiological disturbance and physical factors: distortion idea of body image and weight, and pathological eating behaviors. The Mortality of EDs is 10-15% (2/3 physical complications and 1/3 suicide) [1]. ED includes anorexia nervosa (AN), bulimia nervosa (BN), binge eating disorders and so on. Importantly, the women with ED may suffer from the gynecological endocrinological conditions (amenorrhea, infertility), such as: women who suffered from binge eating have more risk of suffering from amenorrhea and oligomenorrhea than females who reported no binge eating [2]; AN patients has profound low leptin levels which increase pulsatile LH, leading to ovaries enlarging, levels of estradiol increasing which suggest that treating amonorrhea might be key strategy for AN patients [3]; Leptin level was significantly low in AN patients even compared with underweight controls which support that leptin predict AN independently form low BMI [4]. Moreover, EDs are often regarded analogous to cancer among all psychological disorders because people who suffer from it know exactly that their behavior might lead great harm to their body but they could do nothing to stop them, they felt depressed, feared, struggle and hopeless.

Generally, EDs were traditionally thought as influenced by the social expectation and literature excessive emphasizes the beauty of thinness among culture of upper/ middle class in western countries. However, recent evidence from family and twin studies suggests that genetic variances account for 50-70% factors for AN and 28-83% for BN [5].

5-HTTLPR gene which was widely reported associated with human’s behaviors, psychological conditions, psychological disorders [6-8], which located on the SERT gene promoter as a functional polymorphism because of a 44-bp insertion (5-HTTLPR-L) or deletion (5-HTTLPR-S). The transcriptional activity of S is less than L which might be the reasons of 5HTTLPR playing an important role in humans’ psychological features [9]. 

Past studies showed that 5-HTTLPR might be the most associated gene with the heredity of EDs, but they does not talk about the results systematic by five different gene models, different races and subtypes of EDs . The present study tries to further clear the relationship between 5-HTTLPR polymorphism and EDs systematically.

An electronic search of the literature was performed to identify association studies investigating the influence the 5-HTTLPR polymorphisms on EDs. Pubmed, Medline, PsychINFO of web knowledge, and CNKI (Chinese National Knowledgey Infrastructure) were searched until April 2012. The search terms were “5-HTTLPR”, ”gene”, “Eating disorder”, “Anorexia Nervosa”, and “Bulimia nervosa”. Reference lists from identified articles and reviews were used as well to find additional studied to be included.

Inclusion criteria of studies in the meta-analysis

Studies fulfilled the following criteria were included: a) they were published studies; b) they were case-control studies and the genetic frequencies of the 5-HTTLPR could be got; c) they were written in English or Chinese; d) the participants for studies are unrelated and does not have other special including criteria expect for the diagnosis of EDs. Studies were excluded if: a) no 5-HTTLPR genotypic frequencies were given for patients; b) the participants has other special including criteria; c) the participants are related (for example family members).

Data extraction

The following information was extracted from each study: name of first author, diagnosis for including participants and control group, publication year, country, diagnosis and assessment criteria, number of sample and control group.

In this meta-analysis, one included paper had participants of EDNOS (Bingeing type) and EDNOS (Purging type), while another included paper involved participants of sub-clinical bulimia nervosa. For convenience, we call BN, EDNOS (Bingeing type), EDNOS (Purging type) and sub-clinical BN as BN related in this paper.

Statistical analysis

Five models were used in this study: L allele vs. S allele, LL genotype vs. SS genotype, LS genotype vs. SS genotype, LL genotype vs. SS+SL genotype and LL+LS genotype vs. SS genotype. The present study also talks EDS as a whole, AN and BN related. The studies also divide participants into two groups: western population (Caucasian, European, Poland, Italian, German, Australian) and Asians (Japanese and Chinese) to study about because the researchers noticed that the frequencies of L allele and S allele in those two kinds of population are significantly different. Not all figures are included in this paper because of the limitation of length of paper. A random-effects model was used for analyses where there was significant heterogeneity between studies. Hardy-Weinberg equilibrium was examined in all studies included in the meta-analysis.

Data were entered into the Cochrane Collaboration review manager software (RevMan version 5). Individual and pooled 95% confidence intervals (CIs) were calculated. For studies not including controls, we use the data of control group from other studies which has similar participants (ethnic) with them: The Steiger’s study published in 2008 [10] get control group from his study published in 2009 [11]. Paper of Ribases published in 2008 [12] gets control group data from paper of Lauzurica [13]. Both the two studies were taken in Spain. The data control group in [14] data were applied in three studies Frieling, 2006 [15], Urwin, [16], Wonderlich, 2005 [17].

Study characteristics

the initial search yielded 23 papers and the abstracts were carefully reviewed in order to investigate if they fulfill the inclusion criteria. Of the 23 articles, four were excluded because they were reviews Martásková, 2009 [18], Calati, 2011 [19], Polsinelli, 2012 [20], Castellini, 2012 [21]. One paper [22] Hinney, 1997 was excluded because the participant were all obese populations. One was excluded because the participants were adolescent and young adult female twins and female non-twin siblings Munn-Chernoff, 2012 [23]. After all, 17 articles were included in this study. The information from these studies was summarized in (Table 1).

Meta-analysis

Data from the 17 samples were not all analyzed because some data are absence, for example, in Ribasés’s study of 2008 [12], only the number of LL genotype participants and the total number of all participant could be got from the article? Therefore, that paper only analyzed the model of LL genotype vs. SS+LL genotype.

L Allele vs. S Allele Model

Eating bisorders: In fixed effect model, I² =81%, indicating high heterogeneity, random-effects model was selected. However, no significant association was found (Z=1.33, p=0.19).

Anorexia nervosa: We include 10 papers [14-16], [21], [24- 29] (In following 4 models about AN we include the same 10 papers)

We use random-effects model: No significant association was (Z = 1.55, P = 0.12),

Heterogeneity was high: Tau² = 0.14; Chi² = 38.66, df = 9 (P < 0.001); I² = 77%

BN Related: We include 8 papers [10,13,15,17,25,28,30,31], (In following 4 models about BN, we include the same papers).

No significant association was found (Z=1.35, p=0.18). There was high between –study heterogeneity (Tau² = 0.23; Chi² = 45.70, df = 7 (P<0.001); I² = 85%)

LL VS. LS+SS Model

Eating disorders: We use random effect model. No significant association was found (Z=0.70, p=0.48). There was high between –study heterogeneity (Tau² = 0.38; Chi² = 72.28, df = 7 (P<0.001); I ² = 78%)

Anorexia nervosa: We use random-effect model. No significant association was found (Z=1.04, p=0.30). There was high between –study heterogeneity (Tau² =0.24; Chi² =24.71, df =9 (P=0.001); I² = 67%)

BN Related: We use random-effect model. No significant association was found (Z=0.60, p=0.55). There was high between –study heterogeneity (Tau² =0.63; Chi² =39.90, df = 6 (P<0.001); I ² =85 %)

LL+LS vs. SS Model

Eating disorders: We use random-model. No significant association was found (Z=1.79, p=0.07). There was high between –study heterogeneity (Tau² =0.46; Chi² =84.42, df =15 (P<0.00001); I² = 82%). P is extremely close to the 0.05, it suggest that LL+LL might be the protector factor for EDs. When delete some article from the list (one or some of the [16,24,30,31]), the P does down below 0.05, suggest there was significant association between L carriers and people who has less possibility to suffer from EDs. So including L in the genotype might be the protector factor in EDs.

Anorexia nervosa: We use random-effect model. There was nearly significant association was found (Z=1.86, p=0.06). There was middle between –study heterogeneity (Tau² =0.01; Chi² =28.53, df =9 (P=0.0008); I² =68 %).

When delete one or all of the articles ([16,24]) from the list, the P goes down below 0.05.

BN related: We use random-effect model. No significant association was found (Z=0.79, p=0.43). There was high between –study heterogeneity (Tau² =0.59; Chi² =30.24, df = 5 (P<0.00001); I ² =83 %)

LL vs. SS Model

Eating disorders: We use random-effect model. No significant association was found (Z=1.10, p=0.27). There was high between –study heterogeneity (Tau² =0.66; Chi² =74.38, df =15 (P<0.00001); I² =80 %)

Delete four articles: Delete ([16,24,30,31]) from the list the p becomes 0.001. It suggested that people whose genetic type is LL has significantly less possibility to suffer from EDs. (Figure 1).

Anorexia nervosa: We use random-effect model. No significant association was found (Z=1.50, p=0.13). There was high between –study heterogeneity (Tau² =0.44; Chi² =28.63, df =9 (P=0.0007); I² =69 %)

Delete two articles: When delete ([16,24]) from the list, the P becomes 0.0003 which suggest that persons whose 5-HTTLPT genetic type is LL has significantly less possibility to suffer from the AN.

BN related: We use random-effect model. No significant association was found (Z=1.05, p=0.29). There was high between –study heterogeneity (Tau² =1.76; Chi² = 49.83, df = 5 (P<0.00001); I² =90 %)

Delete two articles: When delete ([30,31]) from the list, the P became 0.001 and suggested that people of LL gene have significantly possibility to suffer from BN.

LS vs.SS Model

Eating disorders: We use random-effect model. No significant association was found (Z=0.21, p=0.84). There was high between –study heterogeneity (Tau² =0.52; Chi² =18.54, df =15 (P=0,24); I² = 19%)

AN : We use random-effect model. No significant association was found (Z=1.13, p=0.26). There was high between –study heterogeneity (Tau² =0.19; Chi² =23.27, df = 9 (P=0.006); I² =61 %)

Asian ethnics: For only the three Asian articles: [24,25,27].

We use fixed-effect model. No significant association was found (Z=3.00, p=0.003). There was low between –study heterogeneity (Chi² = 3.31, df =2 (P=0.19); I² =40 %). It suggested that in Asian populations: people have LS genetic type has significantly less possibility to suffer from. (Figure 2)

BN related: We use random-effect model. No significant association was found (Z=0.77, p=0.44). There was high between –study heterogeneity (Tau² =0.23; Chi² =13.60, df =5 (P=0.02); I² =63

Delete two articles: Delete ([30,31]) from the list, the P became 0.001 and suggested that people of LL gene have significantly possibility to suffer from BN. (Figure 3)

Table 1: Basic information of included studies.

(Endnotes)
i. EDs: Eating Disorders
ii. EDE: Eating disorder Examination (EDE: Fairburn and Cooper, 1993)
iii. BN: Bulimia Nervosa
iv. EDNOS: Eating Disorder Not otherwise specified
v. DSM-IV-TR: Diagnostic and Statistical Manual of Mental Disorders-Fourth Edition (Text Revision)
vi. EDE-Q: Eating Disorder Examination Questionnaire
vii. CES-D: The Center for Epidemiologic Studies Depression Scale
viii. BASIS-32: The Behavior and symptoms Identification Scale 
ix. BIS-11: The Barratt Impulsivity Scale-version 11
x. SCID-I: Structured Clinical Interview for DSM-IV Axis-I disorders
xi. SCL-90R: Symptom Checklist90 Revised
xii. AN: Anorexia Nervosa
xiii. EDI-2: Eating Disorder Inventory -2
xiv. SCID-IP: Structured Clinical Interview for DSM-IV Patient Edition (First et al, 1997)
xv. DAPQ: Dimensional Assessment of Personality Pathology Basic Questionnaire (DAPP- BQ)
xvi. IBS: Impulsive Behavior Scale
xvii. EDE-Q4: Eating Disorder Examination Questionnaire 4
xviii. MAST/ AD: Michigam Assessment Screening Test/ Alcohol- Drug
xix. IDS-SR: Inventory for Depressive Symptomatology - Self Report
xx. STAI/SSAI: Spielberger Stait- Trait Anxiety Inventory
xxi. MOCI: Maudsley Obsessive- Compulsive Inventory Table 2:
xxii. SCID-II: Structured Clinical Interview for DSM-IV Axis-II disorders
xxiii. MINI: Mini International Neuropsychiatric Interview
xxiv. TCI-R: Temperament and Character Inventory Revised
xxv. BED: Binge Eating Disorder
xxvi. HAMD: Hamilton Depression Scale
xxvii. HAMA: Hamilton Anxiety Scale

The aims of this meta-analysis were (1) to determine the association between 5-HTTLPR and eating disorders as a whole, AN and BN related (including BN and BED).

Eating disorders are coding as F50 in ICD-10 and are coding from 307.1 to 307.50 as shown in (Table 2) [1]. The human 5-HTTLPR is located on chromosome 17q11.1 to 17q.12. The 5-HTTLPR polymorphism consisted of two forms S variant (short) or L variant (long). S form is associated with a lower transcriptional activity compared with L’s. [32]. So the aim of this meta-analysis was to further clarify the association between the 5-HTTLPR and eating disorders using all dominant model, recessive model and additive models. We also compare the gene frequencies of L and S in different eating disorders. These results are different in different genetic comparing models. In some situations where some articles were detected, the L gene was proved to be the protecting factor for different eating disorders.

Compared with the [19] Meta-analysis, which was published in 2010, this review adds three new articles: [21,27] and [24]. Among these three articles, [27] and [24] were taken in Chinese population, which were different from all previous meta-analysis. The paper of Yue (2012) [27] proved that the L was the protecting factor for BN. Chen (2012) [24] reported that AN might also was related with morbid personality, and personality disorder might increase the severity of patients’ EDs, which allow we make further studies in the future.

In LL+LS Vs SS model, When delete some article froms the list (one or some of [16,24,30,31]), then the P value goes down below 0.05, suggest the P was the protecting factor for the EDs. Why we deleted those four articles? The [30-31] were taken in Italian, which might suggest 5HTTLPR gene plays a different role in Italian populations compared with other ethics. In addition, In the paper of [31], the distribution of the 5HTTLPR genotypes was found to be linked with body mass index, body fat mass values and harm avoidance score. So the body mass index, body fat mass and avoidance score might contribute to the BN together with other factors which factors are not significant in the paper of Monetleone 2006 [31], for example eating habits, social culture, community environment and so on. We suggest further research to be done in Italian to find out the factors stop the significant relationship between 5HTTLPR genes distribution and BN. In addition, further research could be taken in other demographic population to find out whether body mass index, body fat mass values and harm avoidance score are important intermediate variable between 5HTTLPR genotypes distribution and BN. For the paper of Monetleone BED [30] , bing eating disorder has features of recurrent episodes of binge eating without compensatory behaviors which make it different all other eating disorder (significant restrict behavior features), which might make this paper not suitable to be included to the analyze the ED as a whole. In addition, BED patients has higher mean weight, which make the BED populations different from other eating disorder populations in many places, such as, physical fitness, internal secretion, blood glucose, social function, self-confidence and so on. So, further study could be taken to find out what factors affect the BED together with 5HTTLPR genotype distributions. In Urwin 2005 [16] 114 anorexia nervosa patients have both parents, which seems to be normal in general, population but should be abnormal in eating disorder patients. Most eating disorder patients have family problems, which mean that some of ED patients do not have both biological parents and even none of them. So the patients in Urwin should have different family background with patients in other studies. So the family factor interfere the 5HTTLPT genotype distribution affect the occurrence of EDs as an immediate reason. Further, more study could be taken to find out the in different family backgrounds, the relationship between 5HTTLPR genotype distribution and EDs.

In conclusion, most studies have suggested that L allele is the protecting factor for eating disorders. The conclusion of the meta-analytic results was: The results showed that the 5-HTTLPR L polymorphism was associated with significantly less risk of eating disorders (EDs) (including EDs, anorexia nervosa (AN) and Bulimia nervosa (BN)) [LL vs. SS (homozygote model), EDs: odds ratio (OR) =0.53, confidence interval (CI): 0.36-0.78; LL vs. SS, BN: OR=0.25, CI=0.11-0.57; LS vs. SS (heterozygote), AN, Asians: OR= 0.62, 95% confidence interval (CI): 0.46-0.85; LS vs. SS (heterozygote), BN: OR= 0.56, 95% CI: 0.39-0.79].

Some reasons why we get significant statistic results when we delete these four articles (2 about BN: [30-31]) and 2 about AN ([16,24]) have been stated in the beginning of the discuss section (such as the ethics difference, BED’s differences with other eating disorders and different family background in Urwin 2005 [16] paper.

In addition Urwin 2005’s [16] participants are all Australian, which is hard for us to get suitable control group data from other papers.

Chen 2006 [24] use medical students as the control group, which could not present the general population, which might affect the comparing result.

Table 2: Eating disorders classifications in ICD-10.

The advantages of this meta-analysis include that (1) It include the most papers until April 4, 14 among all meta-analysis about eating disorders and 5-HTTLPR (2)It is the first article discuss 3(EDs, AN related, BN related)* 5(L vs. S, LL vs. SS, LS vs. SS, LL vs. SS+SL and LL+LS)* 3(All ethnics, western ethnics and Asians ethnics)=40 tree diagrams, especially we found for Asians, people who have LL genetic type has less possibility to suffer from BN.

Besides the further studies we might consider to take in the above of the paper, because of the limitation of time and ability, there were something left to be clarified in the further: (1) What make BED and BN (purging type) different from other AN in 5-HTTLPR gene distribution (2) We want to talk about the relationship with different features of ED with 5-HTTLPR (3) We also want to talk about why the genetic type in Asians and Caucasians have different BN risk factors? (4) It will be better if we could include more participants (different country, ethnic, and other background) and more random control group. (5) The relationship among 5-HTTLPR, AN and other background data need to be further analyzed to clarify the pathologies of AN.

1. Semple D, Smyth R. Oxford handbook of psychiatry. Oxford University Press. 2013.

2. Algars M, Huang L, Von Holle AF, Peat CM, Thornton LM, Lichtenstein P, et al. Binge eating and menstrual dysfunction. J Psychosom Res. 2014; 76: 19-22.

3. Brown LM, Clegg DJ. Estrogen and leptin regulation of endocrinological features of anorexia nervosa. Neuropsychopharmacology: 2013; 38: 237.

4. Föcker M, Timmesfeld N, Scherag S, Bühren K, Langkamp M, Dempfle A, et al. Screening for anorexia nervosa via measurement of serum leptin levels. J Neural Transm. 2011; 118: 571-578.

5. Lee Y, Lin PY. Association between serotonin transporter gene polymorphism and eating disorders: a meta-analytic study. Int J Eat Disord. 2010; 43: 498-504.

6. Sharpley CF, Palanisamy SK, Glyde NS, Dillingham PW, Agnew LL. An update on the interaction between the serotonin transporter promoter variant (5-HTTLPR), stress and depression, plus an exploration of non-confirming findings. Behavioural brain research. 2014; 273: 89-105.

7. Wang LJ, Chen CK, Hsu HJ, Wu IW, Sun CY, Lee CC. Depression, 5HTTLPR and BDNF Val66Met polymorphisms, and plasma BDNF levels in hemodialysis patients with chronic renal failure. Neuropsychiatr Dis Treat. 2014; 10: 1235-1241.

8. Boll S, Gamer M. 5-HTTLPR modulates the recognition accuracy and exploration of emotional facial expressions. Front Behav Neurosci. 2014; 8: 255.

9. Deckert J, Catalano M, Syagailo YV, Bosi M, Okladnova O, Di Bella D, et al. Excess of high activity monoamine oxidase a gene promoter alleles in female patients with panic disorder. Human molecular genetics. 1999; 8: 621-624.

10. Steiger H, Joober R, Gauvin L, Bruce KR, Richardson J, Israel M, et al. Serotonin-system polymorphisms (5-HTTLPR and -1438G/A) and responses of patients with bulimic syndromes to multimodal treatments. J Clin Psychiatry. 2008; 69: 1565-1571.

11. Steiger H, Richardson J, Schmitz N, Joober R, Israel M, Bruce KR,et al. Association of trait-defined, eating-disorder sub-phenotypes with (biallelic and triallelic) 5HTTLPR variations. J Psychiatr Res. 2009; 43: 1086-1094.

12. Ribasés M, Fernández-Aranda F, Gratacòs M, Mercader JM, Casasnovas C, Núñez A, et al. Contribution of the serotoninergic system to anxious and depressive traits that may be partially responsible for the phenotypical variability of bulimia nervosa. Journal of psychiatric research. 2008; 42: 50-7.

13. Lauzurica N, Hurtado A, Escartí A, Delgado M, Barrios V, Morandé G, Soriano J. Polymorphisms within the promoter and the intron 2 of the serotonin transporter gene in a population of bulimic patients. Neurosci Lett. 2003; 352: 226-230.

14. Fumeron F, Betoulle D, Aubert R, Herbeth B, Siest G, Rigaud D. Association of a functional 5-HT transporter gene polymorphism with anorexia nervosa and food intake. Mol Psychiatry. 2001; 6: 9-10.

15. Frieling H, Römer KD, Wilhelm J, Hillemacher T, Kornhuber J, de Zwaan M, Jacoby GE. Association of catecholamine-O-methyltransferase and 5-HTTLPR genotype with eating disorder-related behavior and attitudes in females with eating disorders. Psychiatr Genet. 2006; 16: 205-208.

16. Urwin RE, Nunn KP. Epistatic interaction between the monoamine oxidase A and serotonin transporter genes in anorexia nervosa. European journal of human genetics. 2005; 13: 370-5.

17. Wonderlich SA, Crosby RD, Joiner T, Peterson CB, Bardone-Cone A, Klein M, Crow S. Personality subtyping and bulimia nervosa: psychopathological and genetic correlates. Psychol Med. 2005; 35: 649-657.

18. Martásková D, Šlachtová L, Kemlink D, Záhoráková D, Papežová H. Short Communication Polymorphisms in Serotonin-Related Genes in Anorexia Nervosa. The First Study in Czech Population and Metaanalyses with Previously Performed Studies. Folia Biologica (Praha). 2009; 55: 192-197.

19. Calati R, De Ronchi D, Bellini M, Serretti A. The 5-HTTLPR polymorphism and eating disorders: a meta-analysis. Int J Eat Disord. 2011; 44: 191-199.

20. Polsinelli GN, Levitan RN, De Luca V. 5-HTTLPR polymorphism in bulimia nervosa: a multiple-model meta-analysis. Psychiatr Genet. 2012; 22: 219-225.

21. Castellini G, Ricca V, Lelli L, Bagnoli S, Lucenteforte E, Faravelli C, et al. Association between serotonin transporter gene polymorphism and eating disorders outcome: A 6-year follow-up study. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics. 2012; 159: 491-500.

22. Hinney A, Barth N, Ziegler a, von Prittwitz S, Hamann a, Hennighausen K, Pirke KM. Serotonin transporter gene-linked polymorphic region: allele distributions in relationship to body weight and in anorexia nervosa. Life Sci. 1997; 61: PL 295-303.

23. Munn-Chernoff MA, McQueen MB, Stetler GL, Haberstick BC, Rhee SH, Sobik LE, et al. Examining associations between disordered eating and serotonin transporter gene polymorphisms. International Journal of Eating Disorders. 2012; 45: 556-561.

24. Chen J. A study of clinical features, serotonin and related gene polymorphisms in patients with anorexia nervosa. [Doctor Dissertation]. In press 2006.

25. Matsushita S, Suzuki K, Murayama M, Nishiguchi N, Hishimoto A, Takeda A, et al. Serotonin transporter regulatory region polymorphism is associated with anorexia nervosa. Am J Med Genet B Neuropsychiatr Genet. 2004; 128: 114-117.

26. Sundaramurthy D1, Pieri LF, Gape H, Markham AF, Campbell DA. Analysis of the serotonin transporter gene linked polymorphism (5-HTTLPR) in anorexia nervosa. Am J Med Genet. 2000; 96: 53-55.

27. Yue L, Jia X, Yu S, Zhang R, Yuan A, Jiang W, et al. Study on association between polymorphisms of 5-HTTLPR gene and anorexia nervosa in Chinese hand population. Journal of Clinical Psychiatry. 2012; 22: 302-305.

28. Di Bella DD, Catalano M, Cavallini MC, Riboldi C, Bellodi L. Serotonin transporter linked polymorphic region in anorexia nervosa and bulimia nervosa. Mol Psychiatry. 2000; 5: 233-234.

29. Rybakowski F, Slopien A, Dmitrzak-Weglarz M, Czerski P, Rajewski A, Hauser J. The 5-HT2A -1438 A/G and 5-HTTLPR polymorphisms and personality dimensions in adolescent anorexia nervosa: association study. Neuropsychobiology. 2006; 53: 33-39.

30. Monteleone P, Tortorella A, Castaldo E, Maj M. Association of a functional serotonin transporter gene polymorphism with binge eating disorder. Am J Med Genet B Neuropsychiatr Genet. 2006; 141B: 7-9.

31. Monteleone P, Santonastaso P, Mauri M, Bellodi L, Erzegovesi S, Fuschino A, et al. Investigation of the serotonin transporter regulatory region polymorphism in bulimia nervosa: relationships to harm avoidance, nutritional parameters, and psychiatric comorbidity. Psychosom Med. 2006; 68: 99-103.

32. Greenberg BD, Tolliver TJ, Huang SJ, Li Q, Bengel D, Murphy DL. Genetic variation in the serotonin transporter promoter region affects serotonin uptake in human blood platelets. Am J Med Genet. 1999; 88: 83-87.