Significant outcomes
• We examined the relationships between the dopamine transporter gene (DAT1) and specific personality traits in patients with major depression (MD) and its clinical subgroups on the basis of sex and early- or late-onset disease.
• The DAT1 polymorphisms were not associated with either novelty seeking (NS) or harm avoidance (HA) in any of the depressed patients or in the sex-based subgroups.
• DAT1 promoter variants (rs2550948 and rs2975226) may influence NS in the early-onset subgroup of depressed patients in the Han Chinese population.
Limitations
• The associations could be observed only in the subgroup analysis, which is of lower power because of smaller sample size.
• Given that our study had a cross-sectional design, the diagnosis of MD might be revised to bipolar disorder, and such a revision would introduce heterogeneity into our patient group.
• The study results may not be representative of all the personality traits in the gene–personality association study.
Introduction
The presence of personality pathologies in patients with MD may predict worse outcomes (Reference Mulder1). Cloninger's biosocial theory of personality hypothesises that NS, HA, and reward dependence (RD) correlate with the dopaminergic, serotonergic, and noradrenergic systems, respectively (Reference Cloninger2). Of these three personality traits, HA has the strongest association with depression. Depressed patients have higher HA scores compared with healthy controls, and HA correlates positively with the severity of depression (Reference Celikel, Kose and Cumurcu3–Reference Hansenne, Reggers, Pinto, Kjiri, Ajamier and Ansseau6). MD patients with a low HA score have a better treatment response, which may further affect the disease outcome (Reference Joffe, Bagby, Levitt, Regan and Parker7). In contrast to HA, the effect of NS on depression is controversial. A previous study showed that MD patients displayed lower NS compared with healthy controls (Reference Hansenne, Pitchot, Gonzalez Moreno, Machurot and Ansseau8). An animal study using a rodent model of depression also implied that low-NS rats were more vulnerable to the effects of chronic stress tests (Reference Stedenfeld, Clinton, Kerman, Akil, Watson and Sved9). However, in a recent epidemiological study, high NS was suggested to increase suicidality in depressive adolescent patients (Reference Csorba, Dinya, Ferencz, Steiner, Bertalan and Zsadon10). Therefore, the role of NS in depressive illness requires confirmation.
Exploring the molecular basis of personality dimensions has been the focus of numerous genetic association studies (Reference Heck, Lieb and Ellgas11–Reference Joyce, Mchugh, Light, Rowe, Miller and Kennedy14). The dopaminergic system has been suggested to play a role not only in personality pathologies but also in depression (Reference Nutt15). The dopamine transporter is one of the key factors regulating the cerebral dopaminergic system by mediating the active reuptake of dopamine from the synaptic termini (Reference Giros and Caron16). The human dopamine transporter gene DAT1 (SLC6A3) encompasses 15 exons spanning ∼53 kb and is located on chromosome 5p15.3. It has been shown to influence several psychiatric disorders, including schizophrenia (Reference Huang, Chen and Ma17), bipolar disorder (Reference Mick, Kim and Biederman18), and attention deficit hyperactivity disorder (Reference Sharp, Mcquillin and Gurling19). A variable number of tandem repeats (VNTR) in the 3′-untranslated region (UTR) and several single-nucleotide polymorphisms (SNPs) located in the 5′-UTR promoter region of DAT1 are regarded as functional polymorphisms that alter the expression and function of the dopamine transporter (Reference Vanness, Owens and Kilts20–Reference Kelada, Costa-Mallen and Checkoway25). Because both NS and HA are considered to be associated with the dopaminergic system (Reference Suhara, Yasuno and Sudo26–Reference Yasuno, Suhara and Sudo28), the DAT1 polymorphisms in the 3′-UTR and the promoter region are good candidate variants in genetic linkage and association studies of these personality traits.
However, there have been few studies that have investigated the relationship between DAT1 and personality traits (Reference Reif and Lesch12), and most have examined only the 3′-VNTR polymorphisms to investigate the role of DAT1 in different personality traits and different populations. These studies have shown no evidence of an association between the 3′-VNTR of DAT1 and specific personality traits in alcoholic patients (Reference Anghelescu, Klawe and Singer29), bipolar patients (Reference Savitz, van der Merwe and Ramesar30), and healthy volunteers (Reference Heck, Lieb and Ellgas11). Although the role of DAT1 in patients with MD has been studied (Reference Manki, Kanba and Muramatsu31–Reference Huang, Lu, Shih, Yen and Huang33), the relationship between DAT1 variants and specific personalities has been discussed very rarely in patients with MD and has never been studied in the clinical subgroups of MD. Therefore, the aim of the present study was to explore the relationships between DAT1 variants and the personality traits of patients with MD and its clinical subgroups.
Materials and methods
This study was approved by the Institutional Review Board for the Protection of Human Subjects at the Tri-Service General Hospital, a medical teaching hospital affiliated with the National Defense Medical Center in Taipei, Taiwan. Written informed voluntary consent was obtained from all the participants after they had been given a detailed description of the study procedure. To minimise the effects of ethnic differences on allele frequencies, the study participants were recruited from the Han Chinese population. All participants were unrelated, born in Taiwan, and lived there, and all of their grandparents were of Han Chinese ancestry.
In total, 463 patients with MD were recruited from either outpatient or inpatient clinical settings at the Tri-Service General Hospital. Each patient was first interviewed by a psychiatrist for an initial evaluation and subsequently by a well-trained psychologist using a Chinese version of the Modified Schedule of Affective Disorder and Schizophrenia–Life Time (SADS–L) (Reference Endicott and Spitzer34,Reference Merikangas, Stevens and Fenton35). The diagnosis of MD in each patient was discussed and made on the basis of a consensus opinion according to the criteria of the Diagnostic and Statistical Manual of Mental Disorders, 4th edn. The inter-rater reliability scores for the various psychiatric disorders assessed with SADS–L were good to excellent (e.g. the k value for MD was 0.79) (Reference Huang, Lin and Ko36). Patients with a history of substance or alcohol dependence, organic brain disease, severe medical illness, or any concomitant major psychiatric disorder were excluded from the study. A Chinese version of the Tridimensional Personality Questionnaire (TPQ) was used to assess the personality traits in MD patients when they were in a euthymic state. The TPQ is a 100-item self-report questionnaire that is answered on a dichotomous true/false scale (Reference Chen, Chen, Chen, Yu and Cheng37). We excluded the RD dimension because the inter-rater reliability among the Han Chinese population in Taiwan was low (34 items, Cronbach's α = 0.54) (Reference Chen, Chen, Chen, Yu and Cheng37). Hence, only the NS (32 items, Cronbach's α = 0.70) and HA (34 items, Cronbach's α = 0.87) dimensions were analysed in the present study. The MD patients were also classified into two clinical subgroups on the basis of data from a structured interview: those with early-onset MD (162 subjects) and those with late-onset MD (301 subjects). MD is defined as ‘early onset’ if the first episode of depression occurs before the age of 25 years and as ‘late onset’ if the first episode occurs at the age of 25 years or later, as in a previous genetic study (Reference Marazita, Neiswanger and Cooper38).
After informed consent was obtained from all participants, genomic DNA was extracted from peripheral blood leucocytes using a commercial kit (DNAzol; Invitrogen, Carlsbad, CA, USA). A total of five polymorphisms were selected on the basis of possible functional polymorphisms, including four SNPs (rs2550948 and rs2975226 in the promoter region; rs6347 in exon 9; and rs27072 in exon 15). SNPs of the DAT1 were genotyped using TaqMan assays (Applied Biosystems, Foster City, CA, USA) with FAM™ dye and VIC® dye. All TaqMan probes and primers were purchased from Applied Biosystems. We used the Applied Biosystems StepOne™ software and StepOnePlus™ Real-Time PCR Systems (Applied Biosystems) for thermocycling and data collection.
3′-VNTR of the DAT1 gene was detected using modified polymerase chain reaction (PCR) methods (Reference Huang, Lin and Wan39). PCR was carried out in a total reaction mixture volume of 25 μl containing 50 ng genomic DNA, 20 pmol of each primer, 2 mM of each dNTP, 1.5 mM MgCl2, 50 mM Tris-HCl (pH 8.3 at 25°C), and 1 U of Taq DNA polymerase (Life Technologies, Carlsbad, CA, USA). The cycling protocols carried out on a Perkin Elmer 9700 (Applied Biosystems) included an initial denaturation at 94°C for 5 min, amplification (30 cycles of denaturation at 94°C for 30 s, annealing at 60°C for 30 s, extension at 72°C for 30 s), and a final extension at 72°C for 10 min.
For genotyping accuracy and quality control, we blinded duplicates selected from 50 random samples using two methods: a restriction fragment length polymorphism method and bidirectional direct sequencing with a model 3730 DNA analyser (Applied Biosystems), as described in our previous report (Reference Huang, Chen and Ma17).
We used the independent-sample t-test and Pearson's χ 2 analysis to determine the differences between patient subgroups (male vs. female; early onset vs. late onset) in demographic variables and genotype frequencies. Deviation from Hardy–Weinberg equilibrium (HWE) was calculated for the entire sample using the χ 2-test. Pearson's χ 2-test was followed by Fisher's exact test if the cell sizes were smaller than expected (fewer than five subjects). We combined the minor homozygous group with the heterozygous group of the five polymorphisms to analyse the association between personality traits and the DAT1 polymorphisms. An independent-sample t-test was used to determine whether the different genotypes were associated with the NS and HA scores in patients with MD or its subgroups. Because we analysed 50 genetic markers, we used Bonferroni's correction to adjust for multiple testing and a p value < 0.001 was considered statistically significant. All data were analysed with the statistical software SPSS (version 15.0; SPSS, Taipei, Taiwan). A power analysis was performed using G*power 3.1 software (Reference Buchner, Faul and Erdfelder40). Our total sample size (n = 463) had a power of 0.47 to detect a small effect (effect size = 0.1) and of ∼0.99–1.00 to detect a medium effect (effect size = 0.3) and a large effect (effect size = 0.5) of genotype distributions (χ2 analysis). In terms of personality analyses (t-test for total sample and subgroups), the study had a power of ∼0.35–0.70 to detect a small effect and more than 0.99 to detect medium and large effects; an α level = 0.05 was chosen.
Results
This study included 463 patients with MD (mean age 38.94 ± 14.07 years; 41.0% male). The genotype distributions and allelic frequencies for all the patients with MD and their subgroups are summarized in Table 1. All SNP markers were in HWE (p > 0.05). The genotype frequencies for the 3′-VNTR in the total MD patient sample were 381 (82.3%) for 10/10 repeats, 53 (11.4%) for 9/10 repeats, 18 (3.9%) for 10/11 repeats, seven (1.5%) for 7/10 repeats, three (0.6%) for 6/10 repeats, and one (0.2%) for 9/9 repeats. The major allele was the 10-repeat allele, with a frequency of 91.0%, and the minor allele was the nine-repeat allele, with a frequency of 5.9%, followed by the eleven-, seven-, and six-repeat alleles. The 6/10, 7/10, 9/9, 9/10, and 10/11 repeat genotypes of the 3′-VNTR were combined because they occurred less frequently than the 10/10 repeat genotype. We found a significant difference in the genotype distributions of rs2550948 between the male and female subgroups (p nominal = 0.046) and another significant difference in rs27072 between the early-onset and late-onset subgroups (p nominal = 0.020; see Table 1). However, both p values did not withstand Bonferroni's correction for multiple testing (p corrected = 0.046 × 10 = 0.46 for rs2550948 and p corrected = 0.02 × 10 = 0.2 for rs2975226, respectively).
Table 1 Genotype distributions and allelic frequencies of the polymorphisms in the SLC6A3 (DAT1) gene with MD in a Han Chinese population
MD, major depression; VNTR, variable number tandem repeat.
*Allele 2 is the minor allele, and only alleles with frequency higher than 1% are shown.
†Genotype of 3′-VNTR include 6/10 repeat, 7/10 repeat, 9/9 repeat, 9/10 repeat, and 10/11 repeat were grouped into others.
‡Genotype in MD patients with male subgroup compared with MD patients with female subgroup.
§Genotype in MD patients with early-onset subgroup compared with MD patients with late-onset subgroup.
||Statistical analysis was performed by Fisher's exact test.
We also combined the rare genotypes (frequencies of <10%) for each SNP before analysing the associations between the NS or HA personality traits and the DAT1 polymorphisms. Therefore, the A/A (present in 2.6%) and G/A genotypes of rs2550948, the A/A (present in 2.8%) and T/A genotypes of rs2975226, the G/G (present in 0.9%) and G/A genotypes of rs6347, and the T/T (present in 5.2%) and C/T genotypes of rs27072 were combined. The mean (± standard deviation) NS and HA scores for each genotype of the five polymorphisms in the 463 patients with MD and in the subgroups are shown in Tables 2 and 3, respectively. No differences in mean NS or HA scores were found between the sex-based subgroups, but early-onset MD patients had higher HA scores compared with late-onset MD patients (mean HA for early-onset MD = 21.3 and for late-onset MD = 19.5, p = 0.002; data not shown in the table). We observed no association between the NS or HA personality traits and the DAT1 polymorphisms in the total MD patients or in the female MD patients. However, in the male MD patients, those with the T/T genotype of rs2975226 tended to have lower HA scores compared with patients with the other genotypes (p nominal = 0.003 for rs2975226 in HA; Table 2), but this result did not remain significant after correction for multiple testing. In contrast, we found that those patients with early-onset MD with the G/G genotype of rs2550948 and the T/T genotype of rs2975226 had lower NS scores than did the patients with the other genotype. These associations remained significant after correction for multiple testing (p corrected = 0.001 × 50 = 0.05 for rs2550948 and p corrected = 0.0001 × 50 = 0.005 for rs2975226; Table 3).
Table 2 Association analysis between SLC6A3 (DAT1) gene polymorphisms and specific personality traits in patients with MD
HA, harm avoidance; MD, major depression; NS, novelty seeking; VNTR, variable number tandem repeat.
*Genotype with minor alleles were combined.
†Genotype of 3′-VNTR include 6/10 repeat, 7/10 repeat, 9/9 repeat, 9/10 repeat, and 10/11 repeat were combined into non10R group.
Table 3 Association analysis between SLC6A3 (DAT1) gene polymorphisms and specific personality traits in subgroup of MD patients (early-onset and late-onset)
HA, harm avoidance; MD, major depression; NS, novelty seeking; VNTR, variable number tandem repeat.
*Genotype with minor alleles were combined.
†Genotype of 3′-VNTR include 6/10 repeat, 7/10 repeat, 9/9 repeat, 9/10 repeat, and 10/11 repeat were combined into non10R group.
‡Early-onset here indicates the first episode of depression occurs before age of 25 years and late-onset indicates the first episode of depression occurs in the age of 25 years or later.
Discussion
The study does not support an association between DAT1 polymorphisms and the NS or HA personality traits in any of the MD patients. However, we found that two SNP variants (rs2550948 and rs2975226) were possibly associated with the NS personality in patients with early-onset MD. In the analyses of total MD patients, our results are consistent with those of previous studies that have reported that the DAT1 polymorphisms are not associated with the development of NS or HA in healthy volunteers (Reference Heck, Lieb and Ellgas11), bipolar patients (Reference Savitz, van der Merwe and Ramesar30), or alcoholic patients (Reference Anghelescu, Klawe and Singer29). However, these studies did not consider age-dependent effects, and our positive results implicated that the DAT1 gene may influence specific personality traits in early-onset depressed patients.
Gender differences have been reported in the dopaminergic system in healthy subjects (Reference Shibuya, Kamata, Suzuki, Matsumoto, Goto and Otani13,Reference Mozley, Gur, Mozley and Gur41) and schizophrenic patients (Reference Glenthoj, Mackeprang and Svarer42). In the present study, male MD patients with the A allele of rs2975226 tended to score higher for HA. Shibuya et al. (Reference Shibuya, Kamata, Suzuki, Matsumoto, Goto and Otani13) also postulated that gender difference along with the DAT1 promoter variant might affect specific personality traits. They found that female subjects without the A allele of rs2975226 had lower scores for self-directedness and cooperativeness, but this polymorphism was not associated with either NS or HA in healthy Japanese subjects. Two factors may possibly explain why our results are inconsistent with those of Shibuya et al. (Reference Shibuya, Kamata, Suzuki, Matsumoto, Goto and Otani13). First, different diagnostic groups might be associated with different personality traits. Our study used MD patients to analyse the role of the DAT1 gene in different personality traits, but other studies have investigated healthy subjects (Reference Heck, Lieb and Ellgas11,Reference Shibuya, Kamata, Suzuki, Matsumoto, Goto and Otani13), patients with bipolar disorder (Reference Savitz, van der Merwe and Ramesar30), or alcoholic patients (Reference Anghelescu, Klawe and Singer29). Second, different lifestyles, social status, and environmental stresses might change a subject's vulnerability to various personality traits. For example, in a previous report that compared self-description and depressive symptomatology in Japanese and Caucasian women, the Japanese women were more depressed and had more negative core self-descriptions and less positive self-conceptions than did the Caucasian women, suggesting that a positive self-description is less valued by the Japanese female population (Reference Saint Arnault, Sakamoto and Moriwaki43). Consequently, the relationship between sex, genetic polymorphisms, and disease in populations should be clarified in future studies.
Patients with early-onset MD had greater familial loading, an increased risk for comorbidity and suicidality (Reference Rush44), an increased risk of switching to bipolar disorder (Reference Akiskal, Maser and Zeller45), and poorer outcomes than did patients with late-onset MD (Reference Coryell, Endicott and Keller46). Moreover, the presence of personality pathology is considered one of the significant predictors of remission in early-onset MD patients (Reference Agosti, Hellerstein and Stewart47). Zubenko et al. (Reference Zubenko, Zubenko, Spiker, Giles and Kaplan48) conducted a large series of studies with early-onset (aged <25 years) and recurrent (two or more depressive episodes) MD subgroups. They reported that these homogeneous subgroups displayed greater genetic loading and more disease transmission to their relatives compared with the general population and therefore considered them a good candidate population for family linkage studies. In this study, our results showed significant differences insofar as patients with the A alleles of SNPs rs2550948 and rs2975226 tended to have higher NS scores in early-onset MD patients. However, the effect of NS on depression is still controversial, especially in young subjects. Jylha et al. (Reference Jylha, Mantere and Melartin49) compared the personality traits of patients with bipolar disorder or MD using Cloninger's model. They found that patients with bipolar disorder displayed more NS than did MD patients. It is possible that MD patients with strong NS traits might be at a greater risk for a revision of their diagnosis to bipolar disorder. A recent epidemiological survey implied that high NS was related to suicidality in adolescent MD patients in a Hungarian sample (Reference Csorba, Dinya, Ferencz, Steiner, Bertalan and Zsadon10). Hence, we propose that variants in the DAT1 promoter region may affect NS in early-onset MD patients, but the effect of this phenomenon on disease outcomes is unclear. Early-onset MD patients were implied to have higher HA compared with controls or those with late onset (Reference Smith, Duffy, Stewart, Muir and Blackwood50,Reference Ongur, Farabaugh, Iosifescu, Perlis and Fava51). The study result does not support our hypothesis that the DAT1 variant is functional to affect HA in early-onset MD patients. Other possible genetic risk factors should be investigated in the future.
This study had several limitations. First, although power analysis revealed sufficient power to detect differences in the total sample of MD patients, there may have been substantially less power to detect associations within specific subgroups (e.g. sex specificity or disease-onset specificity). Second, MD is a multifactorial disorder and is often revised to bipolar I or II disorder in patients with early-onset disease. However, our subjects were evaluated cross-sectionally and our diagnoses may not have been consistent with their permanent MD diagnoses, because we could not predict these revisions. Such revisions may have introduced further heterogeneity into our patient group. Third, the present study focuses on Cloninger's original biosocial theory with NS and HA, but other personality traits such as persistence, self-directedness, cooperativeness, and self-transcendence were not assessed. Therefore, the study results may not be representative of all the entire personality traits (i.e. self-directedness, etc.) in the gene–personality association study.
Conclusions
Overall, our results support the notion that DAT1 polymorphisms do not influence the personality traits of all MD patients or the sex-based subgroups of these patients. However, we suggest that DAT1 promoter variants may affect specific personality traits in the early-onset subgroup of depressed patients in the Han Chinese population. Further prospective cohort studies are required to verify our preliminary findings and to confirm the effects of personality susceptibility on long-term disease outcomes.
