Gene–environment interaction analysis of serotonin system markers with adolescent depression

Abstract

We report analyses from a study of gene-environment interaction in adolescent depression. The sample was selected from 1990 adolescents aged 10-20 years: those with depression symptoms in the top or bottom 15% were identified and divided into high or low environmental risk groups. DNA was obtained from 377 adolescents, representing the four quadrants of high or low depression and high or low environmental risk. Markers within, or close to, each of the serotonergic genes 5HTT, HTR2A, HTR2C, MAOA (monoamine oxidase type A) and tryptophan hydroxylase (TPH) were genotyped. Environmental risk group was a nonsignificant predictor and sex was a significant predictor of the depression group. HTR2A and TPH significantly predicted the depression group, independent of the effects of sex, environmental risk group and their interaction. In addition, there was a trend for an effect of 5HTTLPR, which was significant in female subjects. Furthermore, there was a significant genotype-environmental risk interaction for 5HTTLPR in female subjects only, with the effect being in the same direction as another recent study, reaffirming that an important source of genetic heterogeneity is exposure to environmental risk. Molecular Psychiatry (2004) 9, 908-915.

The moderation by the serotonin transporter gene of environmental adversity in the aetiology of mental illness: review and methodological analysis

“…31 The two adolescent studies both show an interaction in the expected direction among females only and trend for an opposite effect in males. 40,42 The G Â E interaction is consistently positive in young adults. 11,15,29,33,35 Note that in the study of Mandelli and colleagues 33 the middle-aged participants are retrospectively interviewed about events preceding their first affective episode, which occurred on average in their early 30s.…”

“…30,31,34,37,39,41,45 One study used a combined cohort/case-control design, selecting participants with extreme scores on either environmental adversity or depression for genotyping. 40 One study examined individuals exposed to a major stressor (myocardial infarction). 38 Two studies used casesonly design to investigate the association of SLE preceding the onset of depression with genotype 49 or the interaction of SLE during the course of the disorder with genotype on current symptom severity.…”

“…The weak or non-significant associations between SLEs and depression in these studies suggest problems with the validity of self-report measures of SLE 50 and, in the study by Gillepie and colleagues 44 possible inaccuracies of the retrospective assessment of depression. 51 Serotonin transporter gene-environment interactionunselected primary care attendees, 37 three studies used convenience samples of students or teachers; 34,35,43 three studies used enriched samples selected based on experience of childhood abuse, 31 high levels of social adversity and depression 40 or extremes on neuroticism 45 and two studies used clinical samples of depressed patients. 33,36 Sample size is a major determinant of power to detect G Â E. 52 It is therefore notable that the two studies with largest samples produced negative findings.…”

“…Taken at face value, there were eleven replications, 15,[29][30][31][32][33][34][35][36][37][38][39] three partial replications in females only, [40][41][42] and three non-replications. [43][44][45] Importantly, two studies with very large samples were negative.…”

“…41 On the other hand, Kendler and colleagues 15 found that a model with recessive s (that is with ll and ls combined) provided best fit to their data and at least one other study showed results consistent with recessive s. 34 Results from child and adolescent samples place heterozygotes intermediate but close to ll homozygotes, suggesting a predominantly recessive s effect for early-onset depression. 30,40,42 When multiple outcomes were considered within one sample, a co-dominant effect was shown for the more common outcome of alcohol abuse, but a recessive s effect was apparent for the less common illicit drug use. 43 In conclusion, the role of 5-HTTLPR heterozygosity in G Â E remains unclear.…”