A new modelling study led by Oxford Population Health identifies which diseases could potentially be avoided in old age through healthy lifestyle choices.
Across the world, populations are ageing but longer lifespans do not necessarily mean more years of good health. Rising burdens of age-associated diseases are placing greater demands on already pressured health services, but is poor health an inevitable part of ageing? A new analysis led by Oxford Population Health has combined mathematical modelling with big data statistics to distinguish between diseases strongly linked to ageing and those that could potentially be avoided by making good choices throughout life. The results have been published today in PNAS Nexus.
Traditional methods (such as proportional Cox hazards models) developed to estimate associations between potential risk factors and disease are often poorly suited to assessing the influence of age, as study lead Dr Anthony Webster (Oxford Population Health) explained: ‘Modern statistical methods were developed to estimate the influence of risk factors, such as smoking, on disease, not to describe how disease risk changes with age. As a result, important age-related patterns of disease have been overlooked. Our study adapted a 1950's "multistage" model of cancer, and used it for the first time to assess the age-related incidence of different diseases.’
According to the multistage model, cancer arises through a sequence of genetic mutations. However, recent studies have discovered that the same genetic mutations that can cause cancer are prevalent in many tissues throughout the body and become increasingly common as we age. Potentially, these may be involved in a range of different diseases besides cancer, making it plausible that the multistage model could also apply to a wide range of other conditions.
To investigate this, the team used data from UK Biobank, a large-scale biomedical database and research resource containing genetic, lifestyle, and health information from half a million UK individuals. Out of 800 different diseases included in the analysis, two variations of the multistage model accurately described the incidence of over 450 diseases, and identified clusters of diseases with very different patterns of incidence.
For the first cluster of diseases (called ‘late-onset’ diseases), the risk of these occurring was negligible throughout most of life, then increased rapidly with old age. These tended to be serious, even fatal, conditions such as stroke, renal failure, aortic valve disorders, cataracts and macular degeneration, and cancers of the blood.
In contrast, the second cluster of diseases (called ‘sporadic’ diseases) had an initially small risk that slowly increased throughout life, with old age having a minimal effect. These included various symptoms such as chest pain, coughs, and bowel problems, along with more specific diseases such as cellulitis (skin infections), tachycardia (fast heart rate), and a range of digestive disorders.
Dr Webster summarised: ‘The rapid increase in the risk of late-onset diseases makes them appear inevitable in old age, whereas the weak influence of ageing on the risk of the sporadic diseases suggests that they might be preventable. If you reach old age without having had a sporadic disease, there appears to be a reasonable statistical chance that you might never experience one.’
For the next stage, the team would like to explore whether there are genetic differences that predispose people to sporadic diseases, and whether other modifiable exposures such as diet or exercise can be used to help avoid them.
Potentially, the model could also help improve early detection of disease. As Dr Webster explained: ‘The model can quantify your risk of different diseases in terms of an effective age. Screening programmes could then be optimised to test individuals with different ages and risk factors, but with an equivalent risk of disease’.