Olanzapine-induced weight gain is associated with the −759C/T and −697G/C polymorphisms of the HTR2C gene

Article metrics


Weight gain, a serious problem associated with some antipsychotic drugs, notably olanzapine and clozapine, was suggested to be associated with −759C/T polymorphism of the 5-HT2C receptor gene. This study aimed to examine a potential association of two functional polymorphisms of the promoter region of this gene: −759C/T (rs3813929) and −697G/C (rs518147), with weight gain after 6 weeks of olanzapine monotherapy. It included 107 patients with schizophrenia; among them 36 are first-episode drug-naïve patients. Analysis was carried out by PCR–restriction fragment length polymorphism. A protective effect of −759T and −697C alleles was found: significantly less patients with −697C (3/51) and no patient with −759T (0/28) alleles experienced body mass index increase 10% (P=0.0006 and 0.002, respectively). The same was true for drug-naïve patients possessing any of the variant alleles. There was a significant association of haplotypes with a 10% body mass index increase (P=0.001). On the basis of the additional statistical analysis, the more important role of −697C allele was suggested.


Weight gain is one of the most important side effects associated with antipsychotic treatment. It adversely influences both somatic health (for example, is associated with diabetes, heart diseases and other serious states) and personal well-being of patients, with a reduction in compliance, and the risk of illness recurrence as a possible consequence. The importance of this problem was shown by Fontaine et al.,1 who found that clozapine, although preventing suicide in 492 per 100 000 schizophrenic patients in the 10-year period, at the same time could be considered responsible for 416 deaths because of the consequences of antipsychotic-induced weight gain.

Among atypical neuroleptics, olanzapine and clozapine have been associated with the highest weight gain, with the mean increase in weight at the level of 5 kg during short-term treatment2 and 6–7 kg during long-term treatment.3

Mechanisms attributed to weight gain are complex, with the involvement of various substances, for example, leptin, insulin, neuropeptide Y or different neurotransmitters. Among neurotransmitters, serotonin has been suggested4 to play an important role in the regulation of feeding behavior in rodents5 and humans,6 in particular through 5-HT2C receptors. It was shown that, compared with wild-type mice, knockout mice lacking this receptor gene were obese due to hyperphagia.7

Genetic factors are thought to play a role in individual response to antipsychotic drugs—both in terms of benefits from the treatment and in susceptibility to side effects. Variations in gene sequence, called polymorphism, might influence the proprieties of encoded proteins, and result in individual response to drugs. Some polymorphisms were earlier found to be associated with weight gain, for example, polymorphisms in ADRA2A or SNAP-25 genes.8 However, the strongest association was shown for polymorphisms in the 5-HT2C receptor gene (HTR2C gene), which supports its role suggested by pharmacological studies.

The HTR2C gene is located on the chromosome Xq24 and contains several polymorphisms. Although the results for an association of the first polymorphism studied, Cys23Ser in the coding region, with antipsychotic-induced weight gain were mostly negative,9, 10 encouraging data were obtained for the polymorphisms in the promoter region of this gene, identified by Yuan et al.,11 namely, −697G/C, −759C/T, −997G/A and (GT)n repeat at −1027 within the 5′ flanking region. A number of studies investigated the potential role of these polymorphisms in weight gain control. Yuan et al.11 hypothesized their association with the obesity status, and specifically of −697C, −759T and −995A alleles with leanness. The role of the −759C/T polymorphism was further examined in the context of neuroleptic-induced weight gain. The −759T variant was found to be associated with lower weight gain in drug-naïve and first-episode psychotic patients,12, 13, 14 multi-episode patients15 and treatment-resistant patients,16 although the results have not been completely consistent17, 18, 19, 20 and a meta-analysis of studies confirmed its role under only one statistical model.21 Another study suggested the role of this polymorphism in resistance to weight loss.22 One study showed the role of this single nucleotide polymorphism (SNP) in a haplotype with the Ser allele of the Cys23Ser and the long allele of the (GT)n polymorphisms.23

The above polymorphisms are of special interest because of their vicinity to the major transcription initiation site in position −703.24, 25 Indeed, a number of studies have proved their functionality; they were shown to influence transcription rates. This may potentially have an effect on response to drugs; as a decrease in serotonin levels increases feeding,16 and this action of serotonin is partly mediated through HTR2C receptor, theoretically a reduction in this receptor number/activity might have a similar effect to low serotonin levels. According to this hypothesis, the variants associated with a higher expression might have a protective effect against weight gain. However, on the basis of the available studies, the direction of changes in expression remains unclear. Yuan et al.11 showed a higher expression of haplotypes containing both −759T and −697C alleles and the −697G allele alone, by 2.58 and 1.44-fold, respectively, when compared with the haplotype containing only wild alleles (−759C and −697G). A higher expression associated with the −759T allele was confirmed by another study,26 although it did not examine the −697G/C SNP. However, Hill et al.27 showed a decrease in HTR2C promoter activity when the above alleles were present. Such a decrease in activity in the presence of −697C allele, with a further reduction by the −759C and the −997G alleles, was also found by McCarthy et al.28 Although the findings are not consistent between the studies, most of them suggested that the role of the −697G/C polymorphism in the regulation of the promoter activity was greater than that of the −759C/T SNP. It is also worth noting that the −759C/T polymorphism seems to be in linkage disequilibrium with the −997G/A SNP. Therefore, the latter one, not the −759C/T, might in fact be the functional one. Yuan et al.11 showed a complete linkage disequilibrium between the above SNPs forming the −997G/−759C and the −997A/−759T haplotypes. However, Buckland et al.26 observed that although the −759T allele confers a higher gene activity, this activity is decreased by the −997A variant allele. More studies on functional effects of those polymorphisms are necessary to clarify the influence of the above SNPs. Also, the mechanisms underlying changes in promoter activity are unknown. One of the possibilities might be alterations in the affinity for transcription factors, another—a disruption of the promoter site, especially in case of the −697 SNP.28 However, this has not been proven yet.

This study aimed to assess whether there is an association between the −697G/C and the −759C/T polymorphisms of the HTR2C gene and olanzapine-induced weight gain. In order to make our study comparable with other studies, we used the criterion of body mass index (BMI) increase 7, 10% and a quantitative analysis with the full range of BMI values (expressed as a % change from baseline values).


As the HTR2C gene is located on the chromosome X, men are hemizygotic (that is, they have only one allele). In this study, 13/53 (24.5%) men had the −759T and 22/53 (41.5%) had the −697C variant alleles. Women presented with all three genotypes for both polymorphisms and the variant alleles were present both in a heterozygotic and a homozygotic form (−759: CT 14/54 (25.9%) and TT 1/54 (1.8%); −697: GC 19/54 (35.2%) and CC 10/54 (18.5%). The results did not deviate from Hardy–Weinberg equilibrium (P>0.05). The linkage disequilibrium values for the two polymorphisms were: D′=1.0, r2=0.32 (Haploview 4.1 Broad Institute, Boston, MA, USA).

In the analysis, we used an approach on the basis of the presence of the variant alleles thought to be protective against weight gain.

Demographic characteristics of patients according to the presence/absence of variant alleles (−759T, −697C) and haplotypes are shown in Table 1. We did not observe any significant difference in age, age at onset, pre-olanzapine treatment duration in earlier treated patients, baseline weight, baseline BMI and baseline psychopathology (PANSS scale, Positive and Negative Syndrome Scale) between these genotype variant groups.

Table 1 Demographic characteristics of the patients according to the presence/absence of the HTR2C gene variant alleles (−759T, −697C) and haplotypes. Data presented as median (minimum–maximum)

By stepwise regression analysis we excluded the influence of sex, age, age at onset and duration of pre-olanzapine antipsychotic treatment on BMI change. Only baseline BMI and baseline total PANSS score had a significant effect on BMI change (P=0.02 and 0.004, respectively).

Analysis of variance was carried out to assess the influence of the alleles when a full range of BMI values, expressed as a % change from baseline BMI, was used. The only statistically significant effect was shown for the −759C allele, which was associated with a higher BMI increase (P=0.038, F=4.41; −759C present: mean 5.89, s.d. 6.25; −759C absent: mean 2.20, s.d. 5.12). However, this association was no longer significant when baseline BMI and baseline total PANSS values were included as cofactors (P=0.187). The results for other variants and haplotypes were not significant (data not shown).

The analysis (χ2-tests) using the criteria of 7 and 10% increase in BMI, considered as indicative of clinically significant weight gain in earlier studies, was carried out. Both examined polymorphisms were separately associated with 10% increase in BMI, also when Bonferroni's correction for multiple testing was applied (P<0.001; Tables 2 and 3). No patient with the −759T allele experienced a 10% BMI increase, whereas it occurred in 26.6% of the patients without this allele (Table 2). Also, only 6.0% of the patients in possession of the −697C allele experienced a 10% increase in BMI compared with 32% of the patients without this allele (Table 3). To exclude the blurring of the results by initial BMI values falling out of normal range, the data were re-analyzed including only 67 patients with the initial BMI within the normal range (18.5–24.9). Association of genotypes with a 10% BMI increase was confirmed in this subgroup: χ2=5.71, P=0.01 for the possession of the −759T allele and χ2=5.97, P=0.01 for the possession of the −697C allele. Again, the associations remained significant after Bonferroni's correction for multiple allele testing was applied.

Table 2 Distribution of the –759C/T polymorphism of the HTR2C gene according to 10% BMI increase (n=number of patients): (A) general distribution of genotypes; (B) Distribution of genotypes according to the presence/absence of the variant –759T allele.
Table 3 Distribution of the –697G/C polymorphism of the HTR2C gene according to 10% BMI increase (n=number of patients): (A) general distribution of genotypes; (B) Distribution of genotypes according to the presence/absence of the variant –697C allele.

Application of the 7% increase in BMI cutoff level did not yield any significant results for any of the examined polymorphisms (P>0.05).

Additionally, an analysis for haplotypes of the −697G/C and the −759C/T polymorphisms was carried out. The association of haplotypes with a 10% BMI increase was found (χ2=13.03, P=0.001). The results are presented in Figure 1. Interestingly, the −759T allele was always accompanied by the −697C allele; however, when the −759T allele was absent, the −697C allele might be either absent or present. The haplotype containing only the −759T allele and lacking the −697C allele was not identified in any of the patients. No patient in possession of both −759T and −697C alleles experienced a BMI increase 10%, whereas it was true for 13% of the patients with only the −697C allele, and 32% of the patients without any of the ‘protective’ alleles. The association persisted after excluding patients with the initial BMI beyond the normal range (P=0.02) (Figure 1).

Figure 1

Distribution of −759C/T/−697G/C haplotypes of the HTR2C gene according to 10% body mass index (BMI) increase. Dots on the chart represent particular patients distributed according to the per cent of a BMI increase. Numbers on the chart represent number of patients with a particular haplotype according to the 10% BMI increase as used in all the analyses. (−) means that the allele is lacking, (+) that the allele is present: −759T(−)/−697C(−) stands for −759C/C+−697G/G haplotype; –759T(−)/−697C(+) stands for −759C/C+−697C/C or −759C/C+−697G/C haplotypes; and −759T(+)/−697C(+) haplotype stands for −759T/T+−697C/C or −759T/T+−697G/C or −759C/T+−697C/C or −759C/T+−697G/C haplotypes. P values are significant after Bonferroni's correction for multiple testing (P<0.025).

The above results were confirmed in the subgroup of 36 drug-naïve first-episode patients, although the significance was lower, and in case of the −759C/T polymorphism we observed threshold values (χ2=3.82, P=0.05 for the −759T allele; χ2=6.97, P=0.008 for the −697C allele; and χ2=6.97, P=0.03 for the haplotypes). In this group no patient possessing either −759T or −697C allele experienced 10% increase in BMI. The association persisted when only patients with a normal initial BMI were included (P=0.06; 0.01 and 0.04, respectively). However, after Bonferroni's correction for multiple allele testing was applied, only association of −697C allele remained significant in this group.

As baseline total PANSS score and BMI values were found by stepwise regression analysis to have a significant effect on BMI change, the additional log-linear analysis was carried out to rule out their possible influence on the postulated association between the weight gain and the presence of the two alleles. The interval variables—BMI and total PANSS score—were transformed to dichotomic variables. The pretreatment 24.9 value for BMI and 103 total PANSS score (median values for total group) were taken as threshold values. The log-linear analysis was carried out with four design variables: the pretreatment BMI (< or 25.0) (variable1) and total PANSS (< or 103) (variable2), the presence/absence of −759T (variable3) and −697C (variable4) allele, and one response variable: the increase in BMI (< 10% or10%) (variable5) after olanzapine treatment. The automatic model fitting option was used. The model with two interactions: variable4 and variable5, and variable4 and variable3, was necessary and sufficient to describe all the analyzed data (the goodness-of-fit: χ2=11.439; df=25; P=0.99). The response variable—increase in BMI—showed significant negative association only with one design variable—the presence of the –697C allele.

Interactions of PANSS and BMI were checked using the 10% increase of BMI as a threshold value, according to which patients were divided into two subgroups. The baseline BMI value in the subgroup with treatment-associated BMI increase 10% was lower (21.9±3.4) than in the subgroup with BMI increase <10% (24.2±4.4) (P=0.055). There was a significant difference (P=0.03) in baseline total PANSS score—but not in positive, negative and general subscales scores separately—between patients with the BMI increase 10% (95.3±13.8) and those with BMI increase <10% (102.5±19). However, no further significant correlation was observed between BMI increase (both in % and in kg) and % change in PANSS score (total score and positive, negative and general subscales scores) when Spearman's rank coefficient and Mann–Whitney U-test were used to compare the above subgroups.


This study reports a protective role of the −759T variant allele of the −759C/T polymorphism (rs3813929) and the −697C variant allele of the −697G/C polymorphism (rs518147) of the HTR2C gene against antipsychotic-induced substantial weight gain defined as 10% increase in BMI. To our knowledge, this was the first study that investigated a role of the −697G/C polymorphism in antipsychotic-induced weight gain. The analysis of haplotypes suggested that the presence of the −697C variant was crucial for the protective effect, and the −759T might have an additive effect.

The HTR2C receptor has recently received attention as a factor possibly influencing weight gain both in the general population and in patients treated with antipsychotic drugs. Currently, a protective role of the −759T allele against weight gain seems to be the most consistent finding. Yuan et al.11 found a higher frequency of this allele in the non-obese group of individuals without schizophrenia. Pooley et al.22 found an association of −759C/T heterozygotes with resistance to weight loss in a group of 120 obese women. Reynolds et al.13 conducted the first study that examined an influence of this polymorphism on weight gain during antipsychotic treatment and found a lower increase in BMI after 6 and 10 weeks of the treatment in drug-naïve patients possessing the −759T allele, with no patient in this group experiencing a BMI increase >7% after 6 weeks. In the first study using olanzapine, an association of the −759T allele with <10% weight gain increase after 6 weeks of the treatment was found in 42 patients with a history of multiple psychotic episodes.15 In another study,16 including 41 treatment resistant patients treated with clozapine, a higher frequency of the −759T allele among patients with BMI increase <7% was observed. The protective effect of the −759T allele against BMI increase after 6 weeks, 3 and 9 months of the treatment (greatest at 6 weeks) was also found in a recent study of 73 drug-naïve patients with first psychotic episode.14 Patients were treated with different antipsychotics, but the results held true for a small subgroup of 19 patients treated with olanzapine. The protective role of this SNP was further supported by Ryu et al.29 in the Korean population. A recent meta-analysis by DeLuca et al.21 found a weak support for −759C/T polymorphism role in weight gain under one statistical model. One study showed the role of the haplotype containing this allele, Ser allele of the Cys23Ser and long allele of the (GT)n polymorphisms.23 However, four other studies failed to show the protective role of the −759T allele: three including 80–97 treatment-resistant patients receiving clozapine, measurements taken after 4 months,17 6 weeks18 and 12 weeks,19 and another study including 79 patients treated with olanzapine for at least 3 months.20 In one study13 the protective effect was shown for men, whereas there were only slight differences in BMI in the subgroup of women stratified by genotype. Although our study confirmed the protective role of the −759T allele against weight gain, we cannot confirm sex-dependent differences reported in the above-cited paper (results not presented).

Studies on the −697G/C polymorphism in the context of weight gain are scarce. Yuan et al.11 found a higher frequency of the −697C (variant) allele in non-obese individuals without schizophrenia.

The only study that examined haplotypes including both above SNPs in the context of obesity was that of Yuan et al.11 In a group of 586 men they observed an association of haplotypes including both the −759T and the −697C alleles with leanness, those not including any of the ‘protective’ alleles with obesity and a tendency of the haplotype containing only the −697C allele to prevail in non-obese individuals. Our results are consistent with those findings. In our group we observed a protective effect against antipsychotic-induced weight gain in case of both haplotypes containing −697C allele. The data (Figure 1) suggest that the major factor in this phenomenon might be the presence of the −697C, and not the −759T allele. The presence of the −697C allele seems to be sufficient to protect most of the patients against weight gain. We hypothesize that the role of the −759T allele adds to this effect, as no patient with both −759T and −697C alleles had a substantial BMI increase, and it was true for a small proportion of patients with only −759T allele but lacking the −697C allele. This hypothesis is supported by statistical analysis showing a stronger association of 10% BMI increase with the presence of the −697C allele (Table 3) than the −759T allele (Table 2). Moreover, the significance attributed to the presence of the −759T allele when the alleles were analyzed separately (Table 2) may be in fact a result of including patients with both ‘protective’ alleles into the group, in which −697C might have a primary effect. The results of log-linear analysis support this hypothesis. However, the number of patients included into the study is too small to draw more definite conclusions. In order to estimate the influence of −759T allele itself it would be helpful to analyze the haplotype in which the −759T allele would be present and the −697C allele—absent. Unfortunately, this haplotype was not identified in the examined group, nor was it present in the large group analyzed by Yuan et al.,11 which suggests it might be rare.

Although the results obtained with analysis of variance, showing a higher BMI increase in case of the −759C allele presence, are not directly supportive of the above effect, neither are they contradictory to it, as the −759C allele may play the opposite role to the −759T allele according to the functionality studies, discussed in greater detail in the introduction.

The findings in our first-episode drug-naïve patients were similar to those in the entire group. No patients in possession of any of the variant alleles suffered from 10% weight gain versus 44% of patients possessing none of them. Again, the group was too small to draw more definite conclusions but it is possible that the effect of the variants, especially of the −697C allele, in drug-naïve patients might be stronger than in the group earlier treated with other drugs.

It is worth noting that for the whole group, as well as for a subgroup of drug-naïve patients, a proposed criterion of 7% BMI increase was inadequate and the difference became significant only when the threshold was increased to 10%. The same threshold as the only significant was also found in another study.15

Although the results of most studies, including ours, seem to point at the role of the above polymorphisms in antipsychotic treatment weight gain, all of them suffer from major problems that have to be taken into account, and which may be, at least in part, responsible for the discrepancies in results. Usually only small numbers of patients—very often <50—were included, which makes the statistical analysis prone to errors. Moreover, the type of patients included was highly variable between the studies, ranging from drug-naïve first-episode to treatment resistant chronic patients; those groups could react differently to drugs. Also, the studies varied in terms of the drugs used, with drugs of different propensities of causing weight gain being used. This raises the question of the comparability of the results.

Unfortunately, our study, as others, is not free from drawbacks. Although the study included the highest so far number of patients treated with one antipsychotic drug in monotherapy, it still suffers from limited group size, which limits the power. Thus the results, although promisingly consistent with the earlier findings, have to be considered tentative. Additionally, it was impossible to assess the exact caloric intake, which could differ as patients were allowed to get additional food, such as sweets, from their families. However, the level of activity and the caloric input from the hospital diet did not differ substantially between patients. Another critical issue is compliance to treatment.30 It was improved by including only inpatients into the study and by a strict collaboration between patients and the personnel (for example, psychoeducation was supplied to all patients). However, olanzapine plasma levels would be a useful indicator of compliance and should be considered in further studies.

In summary, our study not only supports the importance of the −759T variant allele of the HTR2C gene in protection against antipsychotic-induced weight gain, but also suggests the role of another polymorphism of this gene, the −697G/C, in this phenomenon. We hypothesized that the −697C allele may in fact have a bigger effect in protection against weight gain and that the role of −759T allele is additive or secondary. This study was one of the few that used olanzapine, and included the highest number of patients so far treated with one drug, especially with such high potential of inducing weight gain.2, 3, 31 The additional strength of the study is an inclusion of a small group of first-episode drug-naïve patients, in whom the results were confirmed. However, because of the shortcomings discussed above, more studies are necessary to establish more definitively the role of HTR2C gene polymorphisms in the phenomenon of weight gain after treatment with antipsychotic drugs.

Materials and methods

The study included 107 participants (mean age 29.3±10.0) with the diagnosis of schizophrenia according to Diagnostic and Statistical Manual of Mental Disorders, fourth edition, text revision (DSM-IVTR) criteria: 103 with paranoid and 4 with undifferentiated subtypes; 54 women (mean age 32.2±11.8 years) and 53 men (mean age 26.3±6.7 years). All patients were of Caucasian race, of Polish origin; they were recruited and assessed in Lublin, and molecular analysis was carried out in Gdansk. The diagnosis was set by experienced psychiatrists, who underwent training in diagnosis setting according to the DSM-IVTR criteria, PANSS assessment and standardization of measurements. Exclusion criteria from the study were: comorbid mental disorder, addiction or somatic illness; concomitant treatment with other antipsychotic or antidepressant drugs or mood stabilizers; earlier use of atypical antipsychotics, including olanzapine or clozapine associated with the highest weight gain; resistance to earlier drug treatment. All participants were inpatients, put on the same hospital diet, although the exact caloric input was not measured and patients were not controlled for an additional intake of sweets and so on. Exercise level was comparable for most of the patients, as most them at the initial phase of psychosis stayed in hospital. All patients were on olanzapine monotherapy in a dose range 20–25 mg per day. Benzodiazepine use was permitted in the first and second week of the treatment. Eleven patients were occasionally given lorazepam in a dose up to 5 mg per day (five patients once, three—twice and three—three times). No other drugs were permitted. Among 107 patients, 71 suffered from multiple episodes in the past and were earlier treated with various typical drugs. None of those patients was earlier treated with atypical antipsychotics. No patient was treatment resistant to typical drugs (switch to olanzapine was mainly due to side effects, for example, extrapiramidal symptoms). Moreover, the group included 36 drug-naïve first-episode patients (mean age 25.4±6.4). The written consent was obtained from all the patients participating in the study, and all the procedures were approved by the local ethical committee.

Weight was measured at the 1st and 42nd day of olanzapine treatment. Weight gain was defined as 7 and 10% increase in BMI after 6 weeks of olanzapine treatment. The above thresholds were chosen to allow comparisons with other studies.


DNA was extracted from blood samples using the phenol–chloroform technique. Genotyping was carried out by a researcher who was unaware of the clinical status of the patients.

Both polymorphisms: −759C/T (rs3813929) and −697G/C (rs518147), were analyzed from the same DNA fragment, amplified from the gene receptor sequence from −855 to −634, using the following primers described earlier:14 forward primer 5′-IndexTermATCTCCACCATGGGTCTCGC-3′ and reverse primer 5′-IndexTermCAATCTAGCCGCTCCAAAGG-3′ (IBB PAN, Warsaw, Poland). The PCR mix contained 60 ng of DNA, 1x PCR buffer, 2 mM of MgCl2, 200 ìM of each dinucleotide triphosphate, 1 ìM of each primer, 2.5 ìl of 50% glycerol and 1 U of Taq polymerase (Fermentas, Vilnius, Lithuania). PCR conditions were as follows: initial denaturation at 95°C for 7 min., 35 cycles: 94°C for 50 s, 57°C for 45 s, 72°C for 30 s and final elongation at 72°C for 10 min. The product length was 252 bp.

Genotypes for both −759C/T and −697G/C polymorphisms were analyzed using restriction fragment length polymorphism, by digestion of PCR products with, respectively, 1 U of FauI restriction enzyme (New England Biolabs, Ipswich, MA, USA) and 1 U of SsiI restriction enzyme (Fermentas) overnight, according to the manufacturer's instructions. Restriction fragments were analyzed on 7% polyacrylamide gel. The patterns were as follows: −759 polymorphism: C allele 133+85+34 bp, T allele: 167+85 bp; −697G/C polymorphism: G allele: 160+54+27+11 or 126+54+34+27+11 bp, C allele: 160+ 81+11 or 126+81+34+11 bp. Two different band patterns for each allele of the −697G/C polymorphism resulted from the fact that SsiI enzyme also digested the polymorphic sites of −759C/T polymorphism. The above described procedure had an additional benefit of double-checking the pattern for −759C/T found by FauI digestion.

Statistical Analysis

Distribution of genotypes was assessed using χ2-test. Between-group comparisons were carried out using Mann–Whitney U and Kruskal–Wallis tests. analysis of variance was carried out to assess the influence of the alleles when a full range of BMI values was used. Stepwise regression analysis and log-linear analysis were carried out to assess the influence of baseline variables on BMI change. As two polymorphisms were analyzed, we used Bonferroni's correction for multiple testing, and P value <0.025 was considered statistically significant. All analyses were carried out using Statistica 7.1 software (Statsoft, Tulsa, OK, USA).

Conflict of interest

All authors declare no conflict of interest.


  1. 1

    Fontaine KR, Heo M, Harrigan EP, Shear CL, Lakshminarayanan M, Casey DE et al. Estimating the consequences of anti-psychotic induced weight gain on health and mortality rate. Psychiatry Res 2001; 101: 277–288.

  2. 2

    Allison DB, Mentore JL, Heo M, Chandler LP, Cappelleri JC, Infante MC et al. Antipsychotic-induced weight gain: a comprehensive research synthesis. Am J Psychiatry 1999; 156: 1686–1696.

  3. 3

    Wirshing DA, Wirshing WC, Kysar L, Berisford MA, Goldstein D, Pashdag J et al. Novel antipsychotics: comparison of weight gain liabilities. J Clin Psychiatry 1999; 60: 358–363.

  4. 4

    Vickers SP, Easton N, Webster LJ, Wyatt A, Bickerdike MJ, Dourish CT et al. Oral administration of the HTR2Creceptor agonist, mCPP, reduces body weight gain in rats over 28 days as a result of maintained hypophagia. Psychopharmacology (Berl) 2003; 167: 274–280.

  5. 5

    Sargent PA, Sharpley AL, Williams C, Goodall EM, Cowen PJ . HTR2C receptor activation decreases appetite and body weight in obese subjects. Psychopharmacology (Berl) 1997; 133: 309–312.

  6. 6

    Comuzzie AG, Allison DB . The search for human obesity genes. Science 1998; 280: 1374–1377.

  7. 7

    Tecott LH, Sun LM, Akana SF, Strack AM, Lowenstein DH, Dallman MF et al. Eating disorder and epilepsy in mice lacking HTR2C serotonin receptors. Nature 1995; 374: 542–546.

  8. 8

    Arranz MJ, de Leon J . Pharmacogenetics and pharmacogenomics of schizophrenia: a review of last decade of research. Mol Psychiatry 2007; 12: 707–747.

  9. 9

    Lentes KU, Hinney A, Ziegler A, Rosenkranz K, Wurmser H, Barth N et al. Evaluation of a Cys23Ser mutation within the human HTR2C receptor gene: no evidence for an association of the mutant allele with obesity or underweight in children, adolescents and young adults. Life Sci 1997; 61: PL9–P16.

  10. 10

    Rietschel M, Naber D, Fimmers R, Moller HJ, Propping P, Nothen MM . Efficacy and side-effects of clozapine not associated with variation in the HTR2C receptor. Neuroreport 1997; 8: 1999–2003.

  11. 11

    Yuan X, Yamada K, Ishiyama-Shigemoto S, Koyama W, Nonaka K . Identification of polymorphic loci in the promoter region of the serotonin HTR2C receptor gene and their association with obesity and type II diabetes. Diabetologia 2000; 43: 373–376.

  12. 12

    Reynolds GP, Zhang ZJ, Zhang XB . Association of antipsychotic drug-induced weight gain with a HTR2C receptor gene polymorphism. Lancet 2002; 359: 2086–2087.

  13. 13

    Reynolds GP, Zhang Z, Zhang X . Polymorphism of the promoter region of the serotonin 5-HT(2C) receptor gene and clozapine-induced weight gain. Am J Psychiatry 2003; 160: 677–679.

  14. 14

    Templeman LA, Reynolds GP, Arranz B, San L . Polymorphisms of the HTR2C receptor and leptin genes are associated with antipsychotic drug-induced weight gain in Caucasian subjects with a first-episode psychosis. Pharmacogenet Genomics 2005; 15: 195–200.

  15. 15

    Ellingrod VL, Perry PJ, Ringold JC, Lund BC, Bever-Stille K, Fleming F et al. Weight gain associated with the −759C/T polymorphism of the HTR2C receptor and olanzapine. Am J Med Genet B Neuropsychiatr Genet 2005; 134: 76–78.

  16. 16

    Miller DD, Ellingrod VL, Holman TL, Buckley PF, Arndt S . Clozapine-induced weight gain associated with the HTR2C receptor −759C/T polymorphism. Am J Med Genet B Neuropsychiatr Genet 2005; 133: 97–100.

  17. 17

    Basile VS, Masellis M, De Luca V, Meltzer HY, Kennedy JL . 759C/T genetic variation of 5HT(2C) receptor and clozapine-induced weight gain. Lancet 2002; 360: 1790–1791.

  18. 18

    Tsai SJ, Hong CJ, Yu YW, Lin CH . 759C/T genetic variation of 5HT(2C) receptor and clozapine-induced weight gain. Lancet 2002; 360: 1790.

  19. 19

    Theisen FM, Hinney A, Bromel T, Heinzel-Gutenbrunner M, Martin M, Krieg JC et al. Lack of association between the −759C/T polymorphism of the HTR2C receptor gene and clozapine-induced weight gain among German schizophrenic individuals. Psychiatr Genet 2004; 14: 139–142.

  20. 20

    Park YM, Cho JH, Kang SG, Choi JE, Lee SH, Kim L et al. Lack of association between the −759C/T polymorphism of the 5-HT2C receptor gene and olanzapine-induced weight gain among Korean schizophrenic patients. J Clin Pharm Ther 2008; 33: 55–60.

  21. 21

    De Luca V, Mueller DJ, de Bartolomeis A, Kennedy JL . Association of the HTR2C gene and antipsychotic induced weight gain: a meta-analysis. Int J Neuropsychopharmacol 2007; 10: 697–704.

  22. 22

    Pooley EC, Fairburn CG, Cooper Z, Sodhi MS, Cowen PJ, Harrison PJ . A HTR2C receptor promoter polymorphism (HTR2C—759C/T) is associated with obesity in women, and with resistance to weight loss in heterozygotes. Am J Med Genet B Neuropsychiatr Genet 2004; 126: 124–127.

  23. 23

    De Luca V, Müller DJ, Hwang R, Lieberman JA, Volavka J, Meltzer HY et al. HTR2C haplotypes and antipsychotics-induced weight gain: X-linked multimarker analysis. Hum Psychopharmacol 2007; 22: 463–467.

  24. 24

    Shih JC, Zhu Q, Chen K . Determination of transcription initiation sites and promoter activity of the human 5-HT2A receptor gene. Behav Brain Res 1996; 73: 59–62.

  25. 25

    Xie E, Zhu L, Zhao L, Chang LS . The human serotonin 5-HT2C receptor: complete cDNA, genomic structure, and alternatively spliced variant. Genomics 1996; 35: 551–561.

  26. 26

    Buckland PR, Hoogendoorn B, Guy CA, Smith SK, Coleman SL, O'Donovan MC . Low gene expression conferred by association of an allele of the HTR2C receptor gene with antipsychotic-induced weight gain. Am J Psychiatry 2005; 162: 613–615.

  27. 27

    Hill MJ, Reynolds GP . 5-HT2C receptor gene polymorphisms associated with antipsychotic drug action alter promoter activity. Brain Res 2007; 1149: 14–17.

  28. 28

    McCarthy S, Mottagui-Tabar S, Mizuno Y, Sennblad B, Hoffstedt J, Arner P et al. Complex HTR2C linkage disequilibrium and promoter associations with body mass index and serum leptin. Hum Genet 2005; 117: 545–557.

  29. 29

    Ryu S, Cho EY, Park T, Oh S, Jang WS, Kim SK et al. 759 C/T polymorphism of 5-HT2C receptor gene and early phase weight gain associated with antipsychotic drug treatment. Prog Neuropsychopharmacol Biol Psychiatry 2007; 31: 673–677.

  30. 30

    Licinio J, Wong ML . Pharmacogenomics in psychiatry: clinical issues to be considered. Mol Psychiatry 2005; 10: 615.

  31. 31

    Simpson MM, Goetz RR, Devlin MJ, Goetz SA, Walsh BT . Weight gain and antipsychotic medication: differences between antipsychotic-free and treatment periods. J Clin Psychiatry 2001; 62: 694–700.

Download references


This work was financially supported by the Medical University of Gdansk, Poland (Grant W-32).

Author information

Correspondence to B R Godlewska.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Godlewska, B., Olajossy-Hilkesberger, L., Ciwoniuk, M. et al. Olanzapine-induced weight gain is associated with the −759C/T and −697G/C polymorphisms of the HTR2C gene. Pharmacogenomics J 9, 234–241 (2009) doi:10.1038/tpj.2009.18

Download citation


  • olanzapine
  • weight gain
  • HTR2C gene
  • 5-HT2C receptor
  • polymorphism
  • pharmacogenetics

Further reading