A clear empirical link exists between stress and depression in children and adolescents (Vitiello, Reference Vitiello2011), where ‘stress’ refers to environmental adversities that pose challenge for children's development. Stress may occur either as an acute event or a chronic adversity, and as a major life event or as minor accumulated events. Stressful events may be normative (e.g. school transition) or pathological (e.g. being bullied), and may be dependent on, or independent of, an individual's attitude. Nevertheless, individuals vary in their response to stress, and how they respond can affect their future adjusting and emotional well-being. Such response heterogeneity is associated with pre-existing individual differences in temperament, personality, cognition and autonomic physiology, all of which are known to be under genetic influences (Plomin, Reference Plomin2001). In recent years, G × E interaction involving specific gene polymorphisms has been identified both in animals and humans. In these studies, particular focus has been placed on the interaction of a deletion/insertion polymorphism in serotonin transporter gene-linked polymorphic region (5-HTTLPR). The short (“S”) allele in the 5-HTTLPR is associated with lower transcriptional efficiency of the promoter compared with the long (“L”) allele in vitro (Lesch et al. Reference Lesch, Bengel, Heils, Sabol, Greenberg, Petri, Benjamin, Muller, Hamer and Murphy1996) and in vivo, both in adults and in children and adolescents (Nobile et al. Reference Nobile, Begni, Giorda, Frigerio, Marino, Molteni, Ferrarese and Battaglia1999). In 2003, Caspi et al. (Table 1), using a prospective longitudinal design, reported that carriers of the S allele exhibit elevated depressive symptoms, diagnosable depression and suicidality after experiencing stressful life events and childhood maltreatment (Caspi et al. Reference Caspi, Sugden, Moffitt, Taylor, Craig, Harrington, McClay, Mill, Martin, Braithwaite and Poulton2003). This generated evidence for validity of the construct of genetically driven individual differences in stress sensitivity. More recently, a new single nucleotide polymorphism (SNP) called rs25531 has been identified within the repeats of 5-HTTLPR: the L_G variant has a level of serotonin transporter expression comparable to the S allele, and both have lower levels than L_A variant (Hu et al. Reference Hu, Lipsky, Zhu, Akhtar, Taubman, Greenberg, Xu, Arnold, Richter, Kennedy, Murphy and Goldman2006). Even if the functional interpretation of the L_G allele has been questioned, some more recent studies used this classification.

Table 1. Summary of the studies described in this review

ALEQ, Adolescent Life Events Questionnaire; BDI, Beck Depression Inventory; BDI-II, Beck Depression Inventory-II; BIS/BAS, Behavioral Inhibition System/Behavioral Activation System Scales; CAPA, Child and Adolescent Psychiatric Assessment Interview; CAQ, Changes and Adjustments Questionnaire; CASQ, Children's Attributional Style Questionnaire; CBCL, Child Behavior Checklist; CCSQ, Children's Cognitive Style Questionnaire; CDI, Children's Depression Inventory; CIDI, Composite International Diagnostic Interview; CRPBI, Children's Report of Parental Behavior Inventory; CSI, Chronic Stress Interview; CTS, Conflict Tactics Scale; DAS, Dyadic Adjustment Scale; DIS, Diagnostic Interview Schedule; ELSI, Economic Living Standards Index; EPQ, Eysenck Personality Questionnaire Revised; FMSS, Five Minute Speech Sample; GADS, Goldberg Depression and Anxiety Scales; HOME Inventory, Home Observation for Measurement of the Environment Inventory; K-SADS-PL, Schedule for Affective Disorders and Schizophrenia for School-Age Children–Present and Lifetime Version; LES, Life Event Scale; LHC, Life History Calendar; LSI, Life Stress Interview; MASC, Multidimensional Anxiety Scale for Children; MCTS, Modified Conflict Tactics Scale; MEL, Munich Events List; PDS, Pubertal Development Scale; RRS, Revised Rutter Scale; SADS-L, Schedule for Affective Disorders and Schizophrenia–Lifetime Version; SCID, Structured Clinical Interview for DSM-IV; SDQ, Strengths and Difficulties Questionnaire; SMFQ, Short Mood and Feelings Questionnaire; TCI, Temperament and Character Inventory; UCLA-CSI, UCLA Chronic Stress Interview; YSR, Youth Self-Report.

Three meta-analyses (Munafo et al. Reference Munafo, Durrant, Lewis and Flint2009; Risch et al. Reference Risch, Herrell, Lehner, Liang, Eaves, Hoh, Griem, Kovacs, Ott and Merikangas2009; Karg et al. Reference Karg, Burmeister, Shedden and Sen2011) and two consecutive reviews of Uher & McGuffin (Reference Uher and McGuffin2008, Reference Uher and McGuffin2010), which include multiple age-ranges, assessing the moderating role of 5-HTTLPR on the relationship between depression and environmental adversities, have shown mixed results. Nevertheless, they were able to identify important study characteristics that influence study outcome, i.e., the stress assessment method and stressor type. The evidence of genetic moderation was stronger among studies that used objective measures (e.g. ‘family structure’) or interpersonal interviews, while it was attenuated by inaccuracies of retrospective self-report questionnaires. The actual duration of the stressor was another critical point: evidence of moderating effect was greater for chronic stressors. Furthermore, in most of these studies, although data have been collected longitudinally, the association between environmental exposure and outcome has been analyzed cross-sectionally. Therefore, even if time is a crucial factor, both in terms of window vulnerability, and in the cascade of maturational events that lead to the unfolding of depression, little is known about the impact of genetic risk factors and environmental exposure over time.

Only few G × E studies have been conducted with youths using rigorous methods, particularly a prospective design and contextual interview to assess both chronic and episodic stress and their prediction on longitudinal change in depressive symptoms, and only recently data have been analysed using time-sensitive techniques. Overall, most of these studies found some evidence supporting a G × E effects (see Table 1). Most studies confirmed the prominent role of chronic adversities (e.g. chronic family stress) v. episodic stressors (Gibb et al. Reference Gibb, Benas and Grassia2009a, Reference Gibb, Uhrlass, Grassia, Benas and McGeary2009b; Hammen et al. Reference Hammen, Brennan, Keenan-Miller, Hazel and Najman2010; Jenness et al. Reference Jenness, Hankin, Abela, Young and Smolen2011, Table 1), and also the possible interaction between early life chronic trauma and recent stressful life events (Kumsta et al. Reference Kumsta, Stevens, Brookes, Schlotz, Castle, Beckett, Kreppner, Rutter and Sonuga-Barke2010), even if mixed results concerning the risk allele were found: two studies (Chipman et al. Reference Chipman, Jorm, Prior, Sanson, Smart, Tan and Easteal2007; Laucht et al. Reference Laucht, Treutlein, Blomeyer, Buchmann, Schmid, Becker, Zimmermann, Schmidt, Esser, Rietschel and Banaschewski2009) reported that the S allele was associated with reduced risk. Only one study (Fergusson et al. Reference Fergusson, Horwood, Miller and Kennedy2011) did not find any evidence of G × E.

The recent use of time-sensitive analytic technique focused attention on other important issues according to developmental perspective. The first issue was the conceptualization and analysis of environmental stress in idiographic stressors (i.e., increases relative to the child's own average level over time) v. nomothetic stressors (i.e., higher stress exposure relative to the sample). In a community sample of youth aged 9–15 years assessed prospectively every 3 months over 1 year (5 waves of data), lagged hierarchical linear modelling analyses showed that 5-HTTLPR interacted with idiographic stressors, but not nomothetic stressors, to predict prospective elevations in depressive symptoms (Hankin et al. Reference Hankin, Jenness, Abela and Smolen2011). This study's findings suggest that 5-HTTLPR allelic variation may underlie individuals’ stress reactivity to increases in one's typical exposure to stressors.

The second issue concerned the possibilities of different developmental trajectories of symptoms and of different time sensitivity to different environmental adversities. In the same way, also genes could encounter attenuation or potentiation of their effects during adolescence.

Petersen et al. (Reference Petersen, Bates, Goodnight, Dodge, Lansford, Pettit, Latendresse and Dick2012) (Table 1) tested the interaction between 5-HTTLPR and stressful life events on adolescents’ trajectories of anxious/depressed symptoms in 574 adolescents followed from ages 12–17 years. They found an effect of G × E in predicting acceleration of anxious/depressed symptoms only at ages 16 and 17, not at the initial level age 12, growth at age 13, or acceleration at age 14 and 15, thus suggesting that vulnerability to acute life events may be stronger in late than early adolescents. In a 5-year follow-up study in general adolescent population sample Nobile et al. (submitted for publication) found a moderating role of 5-HTTLPR on the effect of family status in determining the presence of internalizing problems, but only during early adolescence, while during late-adolescence socio-economic status seems to play a pivotal role, probably influencing the social conditions surrounding youth. Furthermore, 5-HTTLPR polymorphism was found to play an important role in determining the stability of this psychopathological trait: early-adolescence internalizing problems were revealed to be stronger predictors of internalizing problems in late-adolescence, especially in youths who carry the S-allele, thus confirming theory, that 5-HTTLPR S-carriers are characterized by the stable trait of negative affectivity that is converted to psychopathology only under conditions of stress.

In conclusion, a life-course approach seems ideal for understanding determinants of developmental depression: causation in depression appears to be multifactorial, including interaction between genes and stressful life events, or between early life trauma and later stress in life; timing of onset and remission vary widely, indicating different trajectories of symptoms over time. A more ‘dynamic’ G × E perspective could be useful in study regarding transition period, including the use of different measures to assess stressful events in adolescence and the use of time-sensitive modelling techniques that are able to incorporate multiple interacting factors across time.