Background
Depression is a complex mental disorder, and its pathological mechanism has not yet been elucidated. A number of studies suggest that both genetic and environmental factors are involved in the onset or course of depression [
1,
2]. One environmental factor consistently linked to depression is exposure to stressful life events [
3‐
6]. From the genetic viewpoint, the genes associated with the monoamine neurotransmitter serotonin (5-HT) have been well studied. 5-HT is an important neurotransmitter involved in regulating a number of psychological activities, such as emotion, cognition, and circadian rhythms. Functional impairment of the 5-HT system plays an important role in the pathogenesis of depression [
7,
8]. Among the genes of the 5-HT system, the serotonin transporter (5-HTT) gene has captured the most attention because 5-HTT is involved in the reuptake of serotonin at brain synapses, and it is the target of one of the most effective types of antidepressant medications, the selective serotonin reuptake inhibitors (SSRIs) [
8]. The most researched functional polymorphism of the 5-HTT gene to date is the serotonin transporter-linked polymorphic region (5-HTTLPR). This polymorphism has two common alleles, the long (L) and short (S) alleles, differing by a 44-base pair (bp) insertion/deletion [
9]. The S allele is associated with a lower transcriptional efficiency of the 5-HTT gene promoter compared with the L allele [
10].
Caspi et al. [
11] demonstrated that an interaction between 5-HTTLPR and stressful life events (5-HTTLPR × Stress) predicts depression in adults. Individuals with one or two copies of the S allele exhibited more depressive symptoms and diagnosable depression in relation to stressful life events than individuals homozygous for the L allele. Since then, other studies have been carried out to test the 5-HTTLPR × Stress interaction among adults, but the results of these studies were inconsistent. Some studies demonstrated an interaction [
12,
13], but others failed to replicate these results [
14,
15]. A meta-analysis published by Karg et al. [
16] in 2011 supported that 5-HTTLPR moderates the relationship between stress and depression.
The current study aims to address several major limitations in the past research. First, the studies conducted to date used mainly cross-sectional designs and between-subject analyses. Thus, most studies tested 5-HTTLPR, in conjunction with between-subject differences in levels of stress, on predicting individual differences in depression. The current study addresses this shortcoming by utilizing a multi-wave longitudinal design in which levels of stress and depressive symptoms are assessed at multiple time points during the course of the 24-month follow-up interval. We examined 5-HTTLPR, in conjunction with within-subject fluctuations in levels of stress, on predicting within-subject fluctuations in levels of depressive symptoms. More specifically, we examined whether the slopes of the relationship between negative life events and depressive symptoms within individuals varied across individuals with different 5-HTTLPR genotypes.
Second, most studies have portrayed depression as a qualitatively distinct disorder and have used the incidence and diagnoses of depressive disorder as outcome variables. Nevertheless, several researchers suggested that depression may best be viewed as a quantitative deviation from “normal” affective experience and that it is a dimensional, not a categorical, construct [
17,
18]. Therefore, the current study treats depression as a continuous variable and assesses it by self-report depression measures.
Third, the majority of former studies that tested the 5-HTTLPR × Stress interaction in predicting depression focused predominantly on adult samples. Depressive symptoms and diagnoses increase dramatically at the onset of adolescence, particularly for girls [
19]. Longitudinal studies indicate that a gender difference in depression appears at approximately age 13–15 [
20]. Thus, adolescence is a crucial time to study the pathogenesis of genetic and environmental factors on depression. Moreover, a few studies focusing on youth samples indicated that the 5-HTTLPR × Stress interaction effect was not robust, particularly among boys, and that it may be moderated by gender [
21,
22]. The current study expounds the 5-HTTLPR × Stress interaction effect and gender differences in adolescent samples.
In addition, there has been no related research about the 5-HTTLPR × Stress interaction on depression among Chinese adolescent samples. Due to ethnic differences, it is necessary to test whether the conclusions of studies among other races may apply to Chinese adolescents. The current study was the first to collect Chinese adolescent samples and apply a longitudinal design for examining stress, 5-HTTLPR, and the 5-HTTLPR × Stress interaction in predicting depressive symptoms.
Discussion
The current study showed that both stress and depressive symptoms were highest at the initial assessment (Time 1) and decreased during the follow-up interval. A possible reason for this result is that the students had just entered senior high school when the initial assessment was carried out. This major transition may have involved more competition and academic pressures, causing stress and depressive symptoms to have been at the highest levels. As the students acclimated to senior high school, their stress and depression levels may have decreased. In addition, there may have been a practice effect due to the repeated follow-up assessments. Consistent with former epidemiological studies in large samples of Chinese adolescents in a similar age bracket [
29,
30], the results of this study might indicate that there was no gender differences in depression in Chinese adolescents aged 14 to 18.
No evidence of a 5-HTTLPR main effect was found in either females or males in the current study, in accordance with the original research of Caspi et al. [
11] and other studies in adolescent samples [
22,
31,
32]. However, some studies in adolescents have reported a main effect of 5-HTTLPR (e.g. [
33]). Inconsistent results have been reported in studies of other age brackets [
34,
35]. According to the results of the current study, we suggest that there may be no detectable effect of 5-HTTLPR on depressive symptoms of Chinese adolescents.
The main aim of the current study was to examine whether an interaction between 5-HTTLPR and stress predicts depressive symptoms in adolescents. Results showed that a significant interaction effect emerged in females, but not in males. Our results were mainly consistent with several former studies in adolescents [
36‐
38]. These studies indicated that the interaction between 5-HTTLPR and stress should be less robust among male adolescents. It is possible that the interaction effects of genetic variation and stress on depression are moderated by gender, perhaps due to the influence of gender differences in hormone levels on gene expression, serotonergic function, and neural development. For example, estrogen can stimulate a significant increase in 5-HT
2A binding sites, and it acts as a 5-HT modulator. Estrogen not only increases the number of 5-HT
2A receptor binding sites, but it also increases 5-HT synthesis, uptake, and imipramine binding; it decreases 5-HT
1 receptor binding sites and 5-HT transporter mRNA; and it increases the prolactin response to 5-HT agonists [
39‐
42]. Another explanation is the gender × genotype interaction. In other words, genetic factors may have differential, or perhaps even opposite, effects on responses to stress. Research has shown that females are generally more reactive to the depressive effects of stress than males are, and such effects appear to be amplified by genetic factors [
43]. Wüst et al. [
44] confirmed this effect in a study of the association between 5-HTTLPR and hypothalamic-pituitary-adrenal axis regulation. They found that the 5-HTTLPR SS genotype was associated with increased cortisol awakening responses in females. But in males, subjects with the LL genotype exhibited enhanced awakening responses while the SS subjects showed the lowest awakening responses. Therefore, these results may partially explain the gender differences in the effects of 5-HTTLPR on the relationship between stress and depression. In addition, gender differences may in turn explain some of the inconsistencies in former research findings. Some former studies reported no interaction of stress and 5-HTTLPR (e.g., [
34]), and one reason might be that these studies did not test for gender effects and had mixed gender differences in their studies. Therefore, more research is required to determine how these mechanisms might account for the different effects of 5-HTTLPR on depressive symptoms in males vs. females exposed to stress.
The present study benefited from several strengths including 2-year repeated assessments and a multi-wave design, which increased its statistical power. We adopted an idiographic approach and generated a relatively reliable estimate of each adolescent’s degree of stress reaction thereby minimizing the effects of individual differences in variables. Furthermore, we used a general community sample of adolescents, as opposed to a clinical sample with biases that reduce the generalizability, and we applied accurate statistical tests. In addition, to our knowledge, this is the first report on the interaction between 5-HTTLPR and stress predicting depressive symptoms in a longitudinal study in Chinese adolescents aged 14 to 18.
However, some limitations in the current study provide avenues for future research. First, perhaps the most important limitation was the sample size, which limited the number of participants of each gender and genotype. It is important to examine a larger sample and to replicate these results so that we can be more confident in the conclusions. Second, self-report measures were used to assess negative life events, which may not be as accurate as contextual stress interviews. Further research would benefit from using contextual stress interviews in a multi-wave design. Third, the present study only examined the gene × environment interaction in adolescents who were 14 to 18 years old. Thus, future research is necessary to test whether the findings in the present study can be generalized to younger adolescents or children. Finally, we analyzed one specific type of 5-HTTLPR allele variation. Several other sequence variations and single nucleotide polymorphisms (SNPs) in 5-HTT and other genes have been reported to be associated with depression [
45]. Therefore, future studies should examine other gene × environment or gene × gene × environment interactions.
Conclusions
In conclusion, the current study indicates that there are gender differences in the interaction between 5-HTTLPR and stress that predict depressive symptoms in Chinese adolescents. The association between stress and depressive symptoms is moderated by 5-HTTLPR in female adolescents only. In consideration of the limitation of the small sample size, the conclusions of the current study should be examined in a large sample in future research. If the results can be replicated, they will contribute to exploring genetic and environmental factors in pathological mechanism of depression, and providing definite research evidences for prevention and treatment of depression in Chinese adolescents.
Competing interests
All authors declare that they have no competing interests.
Authors’ contributions
Author YSQ, ZXZ and YJY designed the study and wrote the protocol. Author ZY, CQL, CHY and HCJ took part in follow-up assessments and data collecting. Author MQS, WMC, WYP and CL managed statistical analysis. Author MQS completed the literature searches and the first draft of the manuscript. All authors read and approved the final manuscript.