Individuals and populations, risk and resilience over the life course

Although Morgan and Hutchinson highlight mainly the results from the UK-based AESOP study, the scientific basis on which they call for public health policy is much broader. Evidence implicating racial and ethnic disparities in psychosis exists across a number of nations and marginalized groups (Cantor-Graae & Selten, Reference Cantor-Graae and Selten2005; Veling et al. Reference Veling, Selten, Veen, Laan, Blom and Hoek2006; Bresnahan et al. Reference Bresnahan, Begg, Brown, Schaefer, Sohler, Insel, Vella and Susser2007). Still, obvious gaps remain in our knowledge of the nature and timing of relevant risk and protective factors across the life course in individuals and populations.

As articulated by Geoffrey Rose, the causes of rates are not necessarily the same as the causes of individual cases (Rose, Reference Rose1985). The same premise holds for protective factors – the factors attenuating rates of illness are not necessarily the same as those buffering individual risk. There is a dearth of evidence in the same population groups (e.g. Black Caribbeans) across settings, even though cross-context comparisons, guided by the wider evidence base, would serve to unravel important drivers of population rates and individual risk. Such endeavors require funding and international collaboration, and highlighting these critical gaps informs both an immediate and long-term scientific concern: priority to funding for rigorous and systematic scientific investigations of life-course determinants of disparities in psychosis across contexts.

Social determinants – to which Morgan and Hutchinson rightfully draw attention – require careful consideration, from the conceptualization and measurement of exposure to the meaningful interpretation of findings (March et al. Reference March, Hatch, Morgan, Kirkbride, Bresnahan, Fearon and Susser2008a). For instance, the authors provide the example of parental separation before the age of 16 in the AESOP study. This particular example sets into relief several important aspects of this area of research.

First, social constructs are often conflated. Social adversity differs from socioeconomic disadvantage – the underlying social processes may not be the same – and more research is needed to differentiate the two with respect to psychosis. Parental separation before the age of 16, a measure of social adversity, functions as a crude proxy for socioeconomic disadvantage, and, unlike other outcomes such as cardiovascular disease (Marmot et al. Reference Marmot, Rose, Shipley and Hamilton1978), there is little evidence to support a socioeconomic gradient in psychosis. In a recent analysis of data from a large Israeli cohort, Corcoran and colleagues found no gradient of risk for schizophrenia across the range of socioeconomic status at birth (Corcoran et al. Reference Corcoran, Perrin, Harlap, Deutsch, Fennig, Manor, Nahon, Kimhy, Malaspina and Susser2009). Parental separation before age 16, then, is one dimension of a multidimensional complex of socially adverse circumstances that is not captured by the lone variable employed. Other, more direct individual and area-based measures should be conceptualized and measured individually and in concert to improve estimates of the deleterious effects of social adversity.

Second, there is a need to conceptualize and to measure both risk and protective factors across the life course in etiologic studies of psychosis and other forms of mental illness (Rutter & Tienda, Reference Rutter and Tienda2005; Patel & Goodman, Reference Patel and Goodman2007). As with risk factors, determining the nature and timing of protective social factors – such as ethnic density (Faris & Dunham, Reference Faris and Dunham1939; Boydell et al. Reference Boydell, van Os, McKenzie, Allardyce, Goel, McCreadie and Murray2001; Veling et al. Reference Veling, Susser, van Os, Mackenbach, Selten and Hoek2008) – requires systematic investigation over the life course across populations, settings, from neighborhood to nation, and time (March et al. Reference March, Morgan, Bresnahan, Susser, Morgan, McKenzie and Fearon2008b). Moreover, the meaning of protective factors may differ across groups and contexts, and the protective nature of a factor may only emerge when other circumstances are measured. For instance, parental separation before age 16 could confer protection in the context of abuse or neglect. Thus, what is conceived as a risk factor in one set of circumstances might be protective in another. More comprehensive measures of adversity become critical in attempts to disentangle the complexities and variegated circumstances of social life.

Inequality, health disparities, and social policy

Whether psychosis can be addressed alone by public health interventions remains an open question. Just as there is a need for coordinated policy prioritizing international investigations of risk and protective factors across the life course in migrant and ethnic minority populations, there is a need for policies shaping broad social interventions that impact life experience, opportunities, resources, and thereby health disparities across many outcomes, including psychosis (Marmot & Wilkinson, Reference Marmot and Wilkinson2006). In keeping with the recent report by the WHO Commission on Social Determinants of Health, policy at various levels should have as its objectives improving daily living conditions, tackling the inequitable distribution of power, money and resources, and measuring and understanding the problem (as suggested above) and assessing the impact of action (CSDH, 2008). Combating inequality, whether in the form of structural or interpersonal discrimination, would serve to reduce the impact of the social determinants of psychosis and other health outcomes in these disadvantaged populations. Creative approaches to social policy might target not only migrants and racial and ethnic minorities, but the majority populations that produce and reproduce the pathological power structures and social hierarchies that seem to matter for health disparities (Marmot et al. Reference Marmot2004). We have yet to learn whether reducing population rates and individual risk of psychosis requires broader approaches or can be achieved through targeted interventions. Calls for action, like that of Morgan and Hutchinson, provide a necessary and commendable catalyst.