During adolescence, onset of emotional disorders may be fast (days or weeks) or slow (months or years) depending in part on the nature of social adversities. What the exact negative psychological effects are, and why time to onset varies following exposure to negative circumstances, remain almost entirely unknown. An important assumption is that events and difficulties carry a latent and undesirable psychological construct (such as personal threat or negative impact to the self) that can be inferred from a detailed recall of the social characteristics of the experience. Recent advances in neuroscience have opened up possibilities for characterising in a more direct way the intermediate mental and neural processes responsible for organising behavioural responses to different forms of adversity. Coping with social risks may depend on there being a sequential set of mental processes involving emotion recognition, appraisal of the implications for the self, and initiation of control processes that determine the form of behavioural response.

Determining the causation of common emotional and behavioural disorders in young people requires an extension of current social inquiry interview procedures to characterise the neurocognitive processes through which life events and difficulties exert their effects. It is suggested that this can be achieved by incorporating experimental methods of assessing psychological functions and mapping these to neural systems underpinning these functions at different stages of development. Putative intermediate neurochemical systems will not be discussed, although it is acknowledged that their modulating effects may shape the structure and function of the social and emotional brain over time (Reference Goodyer, Park and NethertonGoodyer et al, 2001).

A tentative theoretical framework for further research is proposed, based on some selected illustrations from findings on affective disorders in young people as well as data from studies on adults. Although primarily focused on affective disorders, the framework has implications for life events research in behavioural syndromes as well.

LIFE EVENTS AS CAUSAL PROCESSES IN PSYCHOPATHOLOGICAL CHANGE

Both long-standing and recent social adversities precede and increase the risk for emotional and behavioural psychopathology during the school-age years (Reference Goodyer, Herbert and TamplinGoodyer et al, 2000; Reference Sandberg, Rutter and PicklesSandberg et al, 2001). Such negative experiences, while independent of the symptoms of illness, may be dependent in part on the young person's own actions. In a prospective study of adolescents (Reference Goodyer, Herbert and TamplinGoodyer et al, 2000), no specific combination of long-term or recent family difficulties, marital problems, life events or temperamental predispositions in the adolescent predicted onset, indicating that these adverse environments (although necessary) are rather non-specific antecedent correlates of disorder. Exposure to acute severe personal disappointment or permanent loss 1 month before onset adds significantly to the liability for subsequent disorder. Disappointments, defined as a failure to meet previously held expectations about a forthcoming social experience, were focused in the main on peer-group activities. Despite this improvement in predicting onset, in nearly 50% of subsequent cases the person was not exposed to such a highly proximal undesirable experience. This suggests that the notion of a causal two-stage model of acute on chronic adversities occurring in adolescents with temperamental difficulties is not necessary in a substantial proportion of first-episode cases.

Martha Rueter and colleagues (Reference Rueter, Scaramella and WallaceRueter et al, 1999) investigated the interdependence between persistent parental disagreements and psychiatric symptoms in 13-year-old adolescents over a period of 4 years, and re-interviewed the young people at 19-20 years of age for assessment of emotional disorders. The direct predictors of disorder at age 19-20 years were higher self-reported depression scores at age 13 years and rising depression scores over the subsequent 3 years, the two being correlated. Interestingly, parental disagreements at each year point did not directly predict onset but were crucial moderators of the change in depression scores over time. Increases in parental disagreements correlated with increases in self-reported depressive symptoms. Thus, chronic parental disagreements appear to be slow but relentless producers of depressive symptoms leading to subsequent later clinical disorder. Whether highly specific proximal life events involving peers would have altered the nature of these relationships is not clear, as these were not independently measured. It may also be that social inquiry methods underrecorded important private events such as intrafamilial physical and sexual abuse, which are particularly likely to be correlated with subsequent psychopathology. Such experiences could be embedded in the broader non-specific category of family discord. Nevertheless, these longitudinal observations suggest an evolving psychosocial process over midadolescence exerting significant effects on onset of anxiety and depressive disorders by late adolescence.

Overall, throughout the main adolescent period of risk of affective disorders there appears to be both a fast and a slow pace of onset for first-episode unipolar major depression. Both types seem likely to arise from the same broad set of family risk environments within which a greater understanding of the effects brought about by physically and sexually abusive events is essential. Fast-onset affective disorder appears to be rapidly induced in many teenagers by a highly toxic and rather specific event involving acute peer-group dysfunction or (less commonly) severe loss. If there is not a rapid effect, it looks as though negative life events contribute as further ‘slow burn’ experiences, increasing the liability to onsets evolving by early adult life. Fast-onset events are not specifically associated with a highly emotional temperament: thus, serious disappointments appear to arise more from the ebb and flow of normal adolescent peer-group relations than individual differences in behavioural style.

What precise pathological psychological effects occur within people following exposure to adverse life experiences remains unclear. Investigating this question requires more direct assessment of the mental functions of adolescents who have experienced different types and patterns of personally negative events and difficulties over time.

THE NATURE OF MENTAL FUNCTIONS

Mentality can be considered within two broad domains: sensitivity and performance. Sensitivity can be defined as functions that are open and responsive to environmental stimuli. These include recognising emotions both in the self and in others; putting oneself in other people's shoes; and recalling experiences that are contextually relevant to the immediate environment, evaluating their meaning and subsequently forming possible response behaviours. Performance can be defined as abilities that are required to carry out task-specific actions. They are crucial for organising thoughts and for the control of behavioural responses, and are commonly referred to as executive functions. These processes allow performance to be optimised in complex situations and include planning, attentional flexibility, inhibiting undesirable behavioural responses, and decision-making ability.

MENTAL COHERENCE TO SOCIAL ADVERSITY

When healthy adolescents are exposed to an undesirable life event or difficulty, a complex series of psychological functions may be brought into play in order to effect an adaptive response. I suggest that this is likely to involve activation of components of both the sensitive and the performing mind. Four interconnected psychological steps are hypothesised:

1. (a) processing the immediate emotion response;

2. (b) evaluating the salience of the experience through appraisal and matching with recalled past experiences;

3. (c) organising a mental strategy for responding to this mnemonic process through parallel activation of (at least) decision-making and behavioural inhibition;

4. (d) synchronising the operation of the aforesaid mental functions to reduce the risk of negative effects on the self.

What seems crucial is that this proposed model for processing social experience operates most effectively when the mind is emotionally ‘hot’. This makes efficient emotion processing a key first step of the overall response to social adversity.

If the findings to date are treated as hypothesis-generating, then the level of psychiatric risk following exposure to undesirable life events and difficulties might be related to one or more of three intermediate cognitive endophenotypes: over-sensitive response to the associated emotional tone of the experience; weak control functions, allowing for a disorganised behavioural response; and failure of overall mental integration such that sensitive and control processes lack synchrony. Who might be most liable to these processing weaknesses? Adolescents at high psychosocial risk of psychopathology are likely candidates, perhaps specifically those with latent negative schemas themselves derived from earlier adverse experience.

Prospective studies need to determine the concurrence and coherence of sensitive and performance mental functions in both ‘cold’ and ‘hot’ cognitive states through the developmental period of adolescence. This is likely to require the incorporation of challenge or demand paradigms in study designs and the relation of these to the nature of both recent and past social adversities. This methodological issue is clinically as well as theoretically relevant, as choice of treatment may depend not only on the clinical features and social environment but also on the characteristics of sensitive and performance mental functions — the psychological endophenotype.

THE NEURAL BASIS OF PSYCHOLOGICAL RESPONSE TO LIFE EVENTS

If impaired processing of life events results in measurable affective—cognitive deficits then there should be parallel disruption in the functional neural networks under-pinning these processes. Thus, in healthy adult volunteers, the induction of dysphoric mood and concurrent thinking about personally salient negative events (i.e. ‘hot’ affective—cognitive processing) result in the differential activation of a definable limbic-cortical loop. Different components of this network may be distinctively responsible for the psychology of emotion recognition, cognitive appraisal and behavioural response to social adversity, as well as synchronising the coherence of this complex process. Recognising the emotional tone of a social experience is associated with activity primarily in the amygdala (Reference Calder, Lawrence and YoungCalder et al, 2001; Reference Thomas, Drevetts and WhalenThomas et al, 2001). Processing affect-related meanings of life events appears to be mediated by the medial prefrontal cortex functioning as the executive component for limbic-cortical activity (Reference Teasdale, Howard and CoxTeasdale et al, 1999). The neural basis for evaluating, organising and consolidating the meaning of environmental stimuli in declarative memory is a function of the hippocampus (Reference EichenbaumEichenbaum, 1999), whereas the orbital prefrontal cortex appears preferentially associated with establishing the degree of difficulty of inductive inference from external stimuli (Reference Goel and DolanGoel & Dolan, 2000).

Although the basic programming of these neural networks is genetic, the fine-tuning most probably occurs through social experience in childhood and adolescence (Reference Paus, Collins and EvansPaus et al, 2001). Animal studies have shown deleterious consequences of social stress on neural structure and function, implicating an effect of the social environment on brain through the physiological consequences of persistent interpersonal difficulties (Reference McEwenMcEwen, 1998). Determining the relative effects of chronic and recent life events and difficulties on the patterning of psychological functions and their related neural structures is a major goal of future developmental research. Such vertically integrated science will provide important clues about the interplay between social experiences, mental processes and their neural substrates (Reference Posne and RothbartPosne & Rothbart, 2000).

The clinical implications of such research could be far-reaching. They include identifying socially at-risk populations by their psychological and neural patterns of response to different life event tasks both off-line (outside of the scanning environment) and on-line (using functional magnetic resonance); understanding the nature of developing neurocognitive systems following exposure to different patterns of social adversity through childhood and adolescence; delineating weaknesses and deficits in different components of the limbic-cortical circuit, thereby opening the opportunity for new targets for treatment; and documenting neurocognitive change in patients following treatment.