Are brain age gaps the key to understanding psychiatric disorders?
Who remembers Dr Kawashima’s Brain Training Nintendo DS game? With puzzles that tested mental mathematical ability, working memory and reading fluency, it calculated your “brain age” based on your performance. Lowering my apparent brain age to as close to my real age as possible was an utmost priority for me in the mid-2000s! But as well as being a fun and addictive game, brain age is an informative research parameter that is used to study how the brain develops in both childhood and adulthood. In particular, your brain age gap, that is the difference between your real age and your brain age, is a useful research tool to determine if your brain is aging at a different rate to your chronological age, which may have implications for a number of disease states.
This brain age gap is exactly what Kaufmann and colleagues set out to investigate in their recent Nature Neuroscience paper. They conducted structural MRI scans of over 35,000 brains of people ages 3 to 96 years and used machine learning tools to determine what the brain looks like at different ages and establish a “normal” trajectory for brain aging. They could then scan the brain of a healthy individual and determine their brain age by comparing its structure to this standard trajectory, finally producing a brain age gap as the difference between their real age and their predicted age. A larger brain age gap would suggest that a person’s brain was aging either faster or slower than a neurotypical brain.
The authors went on to investigate the brain age gap of patients with different neurological conditions. Patients with schizophrenia, attention deficit hyperactivity disorder (ADHD), dementia and major depressive disorder to name a few were scanned and compared to age-matched controls. These conditions are particularly interesting when investigating brain age gaps as they have all been linked to abnormal neurodevelopment or changes in the aging process, resulting in psychological symptoms. These scans revealed that patients with schizophrenia, multiple sclerosis and dementia had the largest brain age gaps, with the brains of these patients appearing much older than their chronological age. These brain age gaps were also functionally relevant, with patients with greater brain age gaps having lower performance on a number of tests of cognitive function and a higher degree of disability.
Researchers also noted that there were regional differences in cerebellar-subcortical regions in dementia and MS, regions associated with movement and balance, as well as enhanced brain age in the temporal lobe of patients with schizophrenia. These differences between neurodegenerative (including dementia and MS) and neurodevelopmental conditions (such as schizophrenia) suggest different aging or pathological processes in different brain regions, potentially relating to why overall increased brain age gaps can manifest into several very different psychological conditions. Surprisingly however, there was no evidence of a negative brain age gap, indicating a neurodevelopmental delay, in children with ADHD or autism spectrum disorder (ASD).
Given there are a number of well characterised genetic contributions to many of these psychological conditions, the authors also investigated whether there was overlap between the heritability of brain age gaps and the polygenic architectures known to contribute to the disorders. To do this, they searched single nucleotide polymorphism (SNP) data from over 20,000 people and performed a genome-wide association study (GWAS) to assess how often these common genetic variants occurred in patients with a given condition compared to healthy age- and sex-matched controls. They found that brain age gaps have significant heritability (meaning they are at least partially determined by a person’s genetics, not that they are directly inherited from parents), with common SNPs explaining 24% of the variance in brain age gap across all individuals. Although this doesn’t sound like a lot, it suggests that almost a quarter of the differences in brain age gaps seen in these common disorders have a genetic influence.
The greatest number of loci (points on the genome) associated with both a brain age gap and a psychological condition were for schizophrenia. Given that schizophrenia is thought to be a neurodevelopmental condition, resulting from an interaction between risk genes and environmental factors such as maternal health and nutrition, socioeconomic status and early life cannabis exposure (Dean & Murray, 2005), it is interesting that there was significant overlap between the genes contributing to brain age gap and the condition itself. This suggests a common molecular mechanism that could be the underlying cause of the symptoms of schizophrenia.
One of the key advantages of this paper over previous studies of brain age gap is the sheer enormity of their sample size. In total, 45,615 people were scanned in this study, from multiple international research centres. This allows for robust analysis of brain age gaps in different conditions, even though some are rare, where single site experiments might struggle to recruit enough people to fully power the study.
However, this approach does have some disadvantages. Although the authors were able to use data from a huge number of individuals and did control for age and sex, there will still have been a large amount of heterogeneity in the patient sample. This could include the effects of disease severity, comorbid conditions, substance abuse or other adverse lifestyle factors. In this way, it is difficult to determine whether a brain age gap in a given condition is due to the disease itself or resulting from other lifestyle factors. This heterogeneity will only have been exacerbated by using different scanners between research centres. On the other hand, the genetic studies only used patients with European heritage, meaning these results cannot be extrapolated to other populations. This paper could therefore have benefited from widening its cohort (despite the added variability this would have produced) to take into account different heritages, as well as employing different imaging methods to further investigate brain age gaps. For example, in this paper structural MRI was used to determine brain age, but future experiments could consider using functional MRI to further link the functional differences resulting from conditions to the cognitive effect of brain age gap.
So where can we take this research? The authors suggest using the methodology they developed to run a longitudinal study, to assess how brain age gap changes through the course of a disease. This could be then turned into an early diagnostic tool to identify patients at risk of developing the conditions based on prodromal evidence of brain age acceleration. However, it may prove difficult to translate the outcomes of these large-scale studies into meaningful diagnoses in individual patients. It would also be interesting to do a more in-depth analysis of tissue-specific brain age gaps to see if more localised changes could be observed in ADHD and ASD that may be missed in more global analyses. Furthermore, it should be investigated whether there are brain age gaps in other developmental conditions, for example global development delay, that could explain lower functional capabilities in some children
One of the critical next steps of this research will be to determine if large brain age gaps in these conditions are causal in the psychological symptoms. The authors themselves admit the data presented in this paper is merely correlational, and that it is unclear whether these brain age discrepancies are a “cause, consequence or a confounder” of the disease. Are these changes driving the symptoms, resulting from the condition, or just a different manifestation of the disorder that complicates identifying its pathogenesis? These important distinctions will need to be made to fully understand the implications of brain age gaps in conditions such as schizophrenia, MS and dementia.
Despite the uncertainties, is there anything we can do to limit our own brain age gap? Well, some studies suggest that regularly testing our brains through puzzles, reading and intellectual stimulation may help us to prolong our peak cognitive performance (Willis et al., 2006). Moreover, living a healthy lifestyle, such as a balanced diet with regular exercise, not smoking and limiting alcohol consumption can greatly reduce the risk of vascular dementia, the most prevalent dementia after Alzheimer’s disease. Due to the number of environmental risk factors for schizophrenia, it will also be useful to improve maternal health and nutrition, as well as ensuring children are well cared for and are not exposed to dangerous substances such as cannabis that are known to have a huge negative impact on their mental health. But until we know exactly what genetic mechanisms underpin brain age gaps, we’d better keep following Dr Kawashima’s advice and keep our brains as active and young as possible!
References
Dean K, Murray RM. (2005) Dialogues Clin Neuroscience., 7(1): 69–80.
Willis SL et al. (2006) JAMA, 296(33): 2805–14.
