PREDICTING MRI ABNORMALITIES WITH LONGITUDINAL DATA OF THE WHITEHALL II SUBSTUDY

With increasing age many people develop memory problems and depression. They are predicted by the same risk factors as heart attacks and strokes. We, therefore, believe that memory problems and depression in later life are found with brain changes typical of blood vessel disease. Typically, the brain white matter, where nerve fibre connections are located, is disturbed in its structure, and the grey matter shrinks, due to a loss of brain cells. Fortunately, the brain has a number of strategies to compensate for such damage. Interestingly, some areas of the brain become more active, as others, often nearby or connected, shrink. Brain plasticity thus seems to help keep the brain functioning, in spite of illness related shrinkage. This explains why brain shrinkage by itself does not help the diagnosis of depression or dementia: there is a large overlap in scan changes between patients and healthy volunteers. Only combining measures of brain damage and the amount of compensatory activity allows us to predict the actual deficit in brain function.
This general idea can be applied to dementia: initially areas in the brain responsible for memory (the hippocampus) start to shrink, memory is impaired, but people can still compensate for this by using strategies, such as memos or diaries. Later in the illness, as other (frontal) parts of the brain are affected, these compensation mechanisms fail, and dementia sets in. Similarly, higher quality fibre connections in people of high risk of depression appear to protect them from actually developing the illness.
We have developed a magnetic resonance scan lasting 45 minutes that allows us to collect information about brain shrinkage, white matter integrity, blood flow, resting and working brain networks. The Whitehall II study has examined 7000 civil servants over the last 25 years at yearly intervals, and has therefore information for all the risk factors mentioned above, as well as other information about social background, exercise and mental activity. This kind of information is usually not available for imaging studies, as people tend not to remember such information accurately. Combining long-term Whitehall II information from 800 people with our sensitive and informative MRI scanning sequence will allow us to examine the connection between risk factors and protective factors and brain changes. Furthermore, it will help us establish the effects of these brain changes on current mental state and performance, answering important questions about the natural history of depression and dementia.

The programme will combine multi-modal imaging and cutting edge analysis of brain structure, brain perfusion, white matter integrity and brain function with a rich longitudinal data set, the Whitehall II cohort. Twenty-five year antecedent vascular and metabolic risk trajectories and morbidity, antecedent levels of physical and mental activity, baseline cognitive performance levels and 15-year slopes of memory decrement over time, history of depressed mood, genotype, and measured resilience will be used to model brain changes in 800 subjects. Hypotheses are predicated on the assumption that the brain responds adaptively to any age or illness-related lesion with compensating functional reorganization and repair that result in the restitution of cognitive and mental function and behaviour. Damage to this ?scaffolding structure?, e.g. by widespread vascular damage to executive brain networks, will lead to decompensation of function, and result in clinical presentation with e.g. dementia or depression. We would thus predict that highly functioning individuals may show structural or functional lesions in e.g. hippocampal networks, associated with an increase in activity in scaffolding (e.g. executive) networks. We further predict that such active protective mechanisms will be dependent on such antecedents as vascular risk related behaviour and (mental) activity, in addition to other factors less amenable to treatment and prevention. The presence of detailed and frequently sampled cohort data in the Whitehall II study allows for a unique prospective analysis of the effects of socio-demographic, physical and behavioural factors on brain integrity, and a powerful study design strategy that makes it possible to compare the two extreme expressions of a clinical feature (e.g. depression with first onset in the 60s versus no depressive symptoms over at least 25 years), by controlling and stratifying for potential confounders, such as the conventional variables age, gender and occupational level, but also crucial mechanistic factors, such as vascular risk. Short of a prospective interventional study this will be the most effective and efficient way of establishing time directed associations between socially important variables and brain structure and function in older age. Considering the crucial importance that the growing group of active over 60 year olds will have in society, and the effect that a small shift from disabled to able people will have in this part of society, this is important research.