Are the Effects of Duration of Untreated Psychosis Socially Mediated?

Abstract

Investigation of possible mechanisms by which longer duration of untreated psychosis (DUP) could influence treatment outcomes has focused primarily on evidence for neurotoxic effects. It is also possible that longer DUP has psychosocial effects, which could mediate its impact on outcomes. The evidence of relevance to such socially toxic effects is reviewed, with particular reference to the possible role of social support. There is no definite evidence for social support as a mediator of the influence of DUP, but further investigation of this issue is warranted.

During the past 2 decades, there has been much interest in the possible benefits of providing earlier intervention for psychotic disorders.^1,2 A major factor in the enthusiasm for early intervention is the hypothesis that earlier intervention will lead to improved treatment outcomes as the result of reduced duration of untreated psychosis (DUP). Although there is still uncertainty about the significance of DUP as an influence on recovery, there appears to be a reasonably robust correlation between DUP and symptomatic and functional outcomes during the first several years of treatment.^3,4

The explanation most often offered for any beneficial effects of reducing DUP is the postulate that untreated psychosis results in changes in the brain that are difficult to reverse and therefore treatment delay is somehow neurotoxic.^5,6 This hypothesis has given rise to a substantial body of research on the relation between DUP and measure of brain structure, function, and (or) neurocognitive functioning.^7–10 Evidence concerning possible neurotoxic effects of untreated psychosis is, however, inconclusive.^11,12

There has been relative neglect of other potential mechanisms through which a longer period of untreated psychosis could compromise recovery. These alternate mechanisms could include social psychological factors, which appear to have implications for recovery from psychosis. While acknowledging that there is no definitive evidence for social mediation of any relations between DUP and treatment outcomes, I would argue that further research examining such relations appears justified.

Mediators are factors (B) that at least partially account for a relation between an antecedent factor (A) and a consequence or outcome (C). There are formal criteria for identifying possible mediators. These include that B should be related to A and C, and that the relation between A and C is eliminated or significantly reduced when both A and B are used in predicting C.¹³ Using this framework, is there any evidence suggesting that social psychological factors may act as a mediator between DUP and treatment outcome?

I will focus my discussion on social support, a construct that is widely used when examining the effect of the social environment on health outcomes. In general, social support refers to the functions performed for a person by others, and include provision of emotional, informational, and instrumental assistance.^14,15

Social Support and Treatment Outcome in Psychosis

Cross-sectional studies have shown relations between perceived social support and self-reported indices of recovery.^16–18 In addition, social support is often cited as a very important contributor to recovery by people experiencing a first-episode psychosis (FEP).^19–21 Longitudinal studies could provide stronger evidence that social support influences treatment outcome. Several prospective studies of patients with FEP have now shown that social support is related to symptomatic outcomes,^22,23 functioning, quality of life, and affect.^24–30

Social Support and Duration of Untreated Psychosis

There are 2 primary reasons for supposing that social support could be related to DUP. This first is evidence that members of a person’s social network, such as family, friends, and acquaintances, often play an important role in treatment seeking when psychosis occurs, suggesting that greater social support could bring about shorter DUP.^31–38 The second is evidence that symptoms of psychosis and (or) associated stigma can have a disruptive effect on social relationships, that is, longer DUP could compromise social support.^39–41 Studies of correlates of formally assessed DUP do provide evidence for an inverse relation with social support.^22,42–50 Some of these studies have relied on measures of whether a person is living alone or with others at the time of illness onset,^47,50,51 which may not fully capture social support while others have used indices more directly reflecting social support or overall functioning of a person’s social network.^43–45,48

Norman et al⁵² used data on pathways to care in an effort to assess whether the relation between DUP and social support primarily reflects the role of family and friends in facilitating treatment. They reasoned that if social support was influencing DUP then the relation between the 2 should be explained by the length of time between the onset of psychosis and family or friends recognizing the need for treatment. There was no evidence for the latter delay explaining the relation between DUP and social support, and the relation was not accounted for by possible confounds, such as symptoms, sex, acuity of onset, or premorbid adjustment. These findings were interpreted as indicating that the inverse relation between social support and DUP is not entirely a result of members of a person’s social network facilitating treatment, and may reflect the effect of untreated psychosis on social relationships.

Is Social Support a Mediator?

There are several possible causal pathways by which social support could be related to both DUP and treatment outcomes. First, it should be acknowledged that both social support and DUP may be correlated with other variables, such as premorbid adjustment, social competence, or negative symptoms, which may be the primary influence(s) on outcomes. Regarding this possibility, note that the relation of social support and treatment outcomes has generally been found to be independent of demographic characteristics and clinical presentation,^{19,22,25–28} although there is some evidence that the relation may be stronger for people with a diagnosis within the schizophrenia spectrum.^25,26 Similarly, the relation between DUP and treatment outcomes is generally independent of potential confounds.^3,4

Highlights

• Aspects of social support are correlated with both DUP and treatment outcomes for people with a first episode of a psychotic disorder.

• There is justification for more investigation of the role of social support as a possible mediator of any relations between DUP and treatment outcome.

• Social support is a complex construct and there needs to be more systematic investigation of the importance of its differing components in predicting treatment outcomes in first-episode psychosis.

Social support could influence DUP, with the latter being the primary determinant of subsequent treatment outcomes. Longer periods of untreated psychosis could reflect or bring about poor social support, with the latter being the primary influence on (or mediator of) treatment outcomes. It is also possible that social support and DUP are correlated factors that make independent contributions to treatment outcome. There have been few reports that simultaneously examine the role of DUP and social support in predicting treatment outcomes for patients with FEP. Alvarez-Jimenez et al¹⁹ assessed social support based on having both parents alive, compared with death or loss of contact with one or both parents, and found this index and DUP made independent contributions to predicting likelihood of repeat episodes of psychosis in patients with FEP. Using a more proximal measure, clinician ratings of quality of social relations at time of presentation for treatment,⁵³ Norman et al^22,27 found both social support and DUP made independent contributions to prediction of hospital admissions and occupational functioning during 3-year follow-up, but, by 5 years, DUP was not predicting occupational functioning but social support (particularly as assessed 1 year after initiation of treatment) was predicting it. These few relevant studies do not provide clear indications of whether any effects of DUP on outcomes are mediated by social support. Future research addressing this issue will need to take into consideration several factors related to the nature of social support.

The Nature of Social Support

Social support is a complex construct and there are many subtleties in the distinctions between and descriptions of the relevant processes.^54–56 There are 3 main discriminations underlying the classification of social support: function, source, and perspective.

The functions of social support are frequently divided into instrumental assistance, informational support, and emotional support.^14,15,57 With reference to recovery from psychotic disorders, instrumental assistance could include tangible assistance, such as help with activities of daily living, help with finances, housing, return to school, and employment. Informational assistance could include providing factual information or advice that helps with appraisal of and coping with challenges, such as acceptance of diagnosis and adherence to treatment. Emotional support includes demonstrations of valuing, respecting, and caring for the person. Clearly, each of these functions could have separable effects on different aspects of treatment outcome. Most of the measures used in research cited earlier do not directly assess these separate functions; indeed, measures such as living with others, compared with alone, parental loss or contact with family, do not necessarily reflect the extent to which any of these functions are being served.

Measures of social support used in research on psychotic disorders tend to focus on support from primary groups, such as family and friends.⁵⁸ Such measures do not necessarily reflect resources available from secondary groups, such as coworkers, casual acquaintances, and peer support groups who can be helpful in supporting mental health.^15,59

A distinction is often made between measures of social support that focus on specific instances of support received, and measures of an individual’s perceptions of the social support available.^55,60,61 The latter is generally referred to as perceived support and it has been contended that perceived social support shows the most reliable relations to physical and mental health.^61–63

The research on social support and treatment outcomes for schizophrenia or psychotic disorders has not provided much evidence related to the relative importance of varying aspects of social support. In a cross-sectional study, Norman et al¹⁷ found that a measure designed to reflect perceptions of being valued by family and acquaintances was more strongly related to self-perceived recovery than were measures of perceived informational or instrumental support. This finding appears consistent with the considerable body of research showing that expressed emotion (which includes indices of criticism and hostility) is directly related to likelihood of relapse and other aspects of treatment outcome,⁵⁸ and more recent research implicating expressions of positive thoughts and feelings from others as being particularly important in predicting social adjustment and likelihood of relapse.^64–66 Long periods of active psychosis will likely have impacts on support received from others. Perhaps changes in some domains will differ from trajectories in other areas. For instance, a person may experience increases in instrumental support but perceive reductions in positive regard from others and other aspects of emotional support. There is evidence that negative emotional climate in families is associated with longer periods of untreated psychosis.^67–69 However, there has been no examination of the role of emotional social support, specifically as a possible mediator of the effects of DUP, on treatment outcome.

The suggestion that aspects of emotional support may be particularly important influences on treatment outcomes for people with psychotic disorders is also consistent with evidence that self-esteem is an important influence on onset and course of psychosis and recovery.^70–74 Positive regard from others has repeatedly been found to be a major predictor of self-esteem and associated mood states in nonclinical populations,^75,76 as well as people with psychotic disorders.⁷⁷

Research on social support as a potential mediator of relation between DUP and treatment outcomes would need to not only carefully assess differing dimensions of social support but also consider likely changes in social support over time.²⁸ Unlike DUP, social support is not a static predictor. The evolution of social support over time is likely to reflect ongoing reciprocal influences between differing domains of social support and recovery. Given that social support may well change during the recovery process, to the extent that any effects of DUP are being mediated by social support, we would expect its predictive power to vary over time.

Conclusion

Efforts to understand the mechanisms by which DUP sometimes predicts treatment outcomes for people with psychotic disorders may benefit from including an examination of social psychological factors, such as social support. Neural and social psychological mechanisms should not be seen as incompatible or competing approaches to understanding the impact of untreated psychosis,^78,79 but more careful examination of social mediators seems justified.