The psychiatric disorders that we are all familiar with (e.g., depression, anxiety, schizophrenia) were primarily defined based on expert opinion. An alternative approach is to determine which symptoms group together by analysing their correlations with one another statistically. When this approach is taken, the way that psychiatric symptoms group together begins to look very different to traditional classification systems. Research in younger samples has demonstrated that psychiatric symptoms can be organised into a set of “transdiagnostic dimensions” that encompass symptoms which cut across traditional diagnostic categories. These transdiagnostic dimensions can essentially be thought of as a new way of classifying mental illness. These dimensions also appear to be organised hierarchically, including a general higher-order dimension (at the top of the hierarchy) that captures a general propensity towards any form of mental illness and several lower-order dimensions (at lower levels of the hierarchy) that capture propensities towards more specific forms of mental illness (e.g., emotional symptoms, behavioural symptoms, psychotic symptoms). These transdiagnostic dimensions have been studied extensively in younger populations and have demonstrated ability to predict several important outcomes (e.g., brain health). However, few studies have examined whether this same structure of mental illness emerges in older adults. Given that these transdiagnostic dimensions are able to predict important outcomes in younger samples, they may be able to predict similarly important outcomes in older adults. As such, it is important to determine whether these same dimensions (including their hierarchical organisation) are maintained across the lifespan. Moreover, there may be specific transdiagnostic dimensions that are unique to older populations (e.g., a neurocognitive dimension that captures symptoms indicative of cognitive impairment), which may hold special utility in this age group. Finally, it is important to establish whether these new psychiatric constructs are consistent with the underlying biology of mental illness (e.g., with genetic evidence). Doing so will support the validity of these new constructs and will also contribute to research aiming to identify biological indicators of mental illness. These indicators can then be used to identify older adults that are likely to suffer from different forms of mental illness and are thus in need of preventative interventions (e.g., therapeutic, pharmacological). The UK Biobank data offers a unique opportunity to investigate this in the general population and to do so at a scale that has not be possible in the past.
